WDIFF

draft-ietf-nfsv4-00.txt --> draft-ietf-nfsv4-01.txt

INTERNET-DRAFT B. Callaghan
Document: draft-ietf-nfsv4-00.txt draft-ietf-nfsv4-01.txt M. Eisler
D. Robinson
R. Thurlow
Sun Microsystems
June
D. Noveck
Network Appliance
September 1999

NFS version 4

Status of this Memo

This document is an Internet-Draft and is in full conformance with
all provisions of Section 10 of RFC2026.

Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF), its areas, and its working groups. Note that
other groups may also distribute working documents as Internet-
Drafts.

Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet- Drafts as reference
material or to cite them other than as "work in progress."

The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt

The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html.

Abstract

NFS version 4 is a distributed file system protocol which owes
heritage to NFS versions 2 [RFC1094] and 3 [RFC1813]. Unlike earlier
versions, NFS version 4 supports traditional file access while
integrating support for file locking and the mount protocol. In
addition, support for strong security (and its negotiation), compound
operations, and internationlization have been added. Of course,
attention has been applied to making NFS version 4 operate well in an
Internet environment.

Expires: December 1999 March 2000 [Page 1]

Draft Protocol Specification NFS version 4 June September 1999

Key Words

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
"SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this
document are to be interpreted as described in RFC 2119.

Expires: December 1999 March 2000 [Page 2]

Draft Protocol Specification NFS version 4 June September 1999

Table of Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 6 7
2. RPC and Security Flavor . . . . . . . . . . . . . . . . . . 7 8
2.1. Ports and Transports . . . . . . . . . . . . . . . . . . . 7 8
2.2. Security Flavors . . . . . . . . . . . . . . . . . . . . . 7 8
2.2.1. Security mechanisms for NFS version 4 . . . . . . . . . 7 8
2.2.1.1. Kerberos V5 as security triple . . . . . . . . . . . . 7 8
2.2.1.2. <another security triple> . . . . . . . . . . . . . . 8 9
2.3. Security Negotiation . . . . . . . . . . . . . . . . . . . 8 9
2.3.1. Security Error . . . . . . . . . . . . . . . . . . . . . 8 10
2.3.2. SECINFO . . . . . . . . . . . . . . . . . . . . . . . . 9 10
3. File handles . . . . . . . . . . . . . . . . . . . . . . . 10 11
3.1. Obtaining the First File Handle . . . . . . . . . . . . 10 11
3.1.1. Root File Handle . . . . . . . . . . . . . . . . . . . 10 11
3.1.2. Public File Handle . . . . . . . . . . . . . . . . . . 11 12
3.2. File Handle Types . . . . . . . . . . . . . . . . . . . 11 12
3.2.1. General Properties of a File Handle . . . . . . . . . 12
3.2.2. Persistent File Handle . . . . . . . . . . . . . . . . 12 13
3.2.3. Volatile File Handle . . . . . . . . . . . . . . . . . 13
3.2.4. One Method of Constructing a Volatile File Handle . . 14 15
3.3. Client Recovery from File Handle Expiration . . . . . . 15
4. Basic Data Types . . . . . . . . . . . . . . . . . . . . . 16 17
5. File Attributes . . . . . . . . . . . . . . . . . . . . . 18 19
5.1. Mandatory Attributes . . . . . . . . . . . . . . . . . . 19 20
5.2. Recommended Attributes . . . . . . . . . . . . . . . . . 19 20
5.3. Named Attributes . . . . . . . . . . . . . . . . . . . . 20
5.4. Mandatory Attributes - Definitions . . . . . . . . . . . 20 22
5.5. Recommended Attributes - Definitions . . . . . . . . . . 22 25
5.6. Interpreting owner and owner_group . . . . . . . . . . . 30
6. Filesystem Migration and Replication . . . . . . . . . . . 31
6.1. Replication . . . . . . . . . . . . . . . . . . . . . . 31
6.2. Migration . . . . . . . . . . . . . . . . . . . . . . . 31
6.3. Interpretation of the fs_locations Attribute . . . . . . 32
6.4. Filehandle Recovery for Migration or Replication . . . . 33
7. NFS Server Namespace . . . . . . . . . . . . . . . . . . . 28
6.1. 34
7.1. Server Exports . . . . . . . . . . . . . . . . . . . . . 28
6.2. 34
7.2. Browsing Exports . . . . . . . . . . . . . . . . . . . . 28
6.3. 34
7.3. Server Pseudo File-System . . . . . . . . . . . . . . . 29
6.4. 35
7.4. Multiple Roots . . . . . . . . . . . . . . . . . . . . . 29
6.5. 35
7.5. Filehandle Volatility . . . . . . . . . . . . . . . . . 29
6.6. 35
7.6. Exported Root . . . . . . . . . . . . . . . . . . . . . 29
6.7. 36
7.7. Mount Point Crossing . . . . . . . . . . . . . . . . . . 30
6.8. 36
7.8. Security Policy and Namespace Presentation . . . . . . . 37
7.9. Summary . . . . . . . . . . . . . . . . . . . . . . . . 30
7. 37
8. File Locking . . . . . . . . . . . . . . . . . . . . . . . 31
7.1. 38
8.1. Definitions . . . . . . . . . . . . . . . . . . . . . . 31
7.2. 38
8.2. Locking . . . . . . . . . . . . . . . . . . . . . . . . 32
7.2.1. 39

Expires: March 2000 [Page 3]

Draft Protocol Specification NFS version 4 September 1999

8.2.1. Client ID . . . . . . . . . . . . . . . . . . . . . . 32
7.2.2. 39
8.2.2. nfs_lockowner and stateid definition Definition . . . . . . . . . 34
7.2.3. 41
8.2.3. Use of the stateid . . . . . . . . . . . . . . . . . . 34
7.2.4. 41
8.2.4. Sequencing of lock requests Lock Requests . . . . . . . . . . . . . 35
7.3. 42
8.3. Blocking locks Locks . . . . . . . . . . . . . . . . . . . . . 35
7.4. 42
8.4. Lease renewal Renewal . . . . . . . . . . . . . . . . . . . . . 36
7.5. 43
8.5. Crash recovery Recovery . . . . . . . . . . . . . . . . . . . . . 36

Expires: December 1999 [Page 3]

Draft Protocol Specification NFS version 4 June 1999

7.6. 43
8.5.1. Client Failure and Recovery . . . . . . . . . . . . . 43
8.5.2. Server revocation of locks Failure and Recovery . . . . . . . . . . . . . 44
8.5.3. Network Partitions and Recovery . . 37
7.7. Share Reservations . . . . . . . . . 44
8.6. Server Revocation of Locks . . . . . . . . . . 38
7.8. OPEN/CLOSE procedures . . . . . 45
8.7. Share Reservations . . . . . . . . . . . . 38
8. Defined Error Numbers . . . . . . . 46
8.8. OPEN/CLOSE Procedures . . . . . . . . . . . 40 . . . . . . 47
9. NFS Version 4 Requests Client-Side Caching . . . . . . . . . . . . . . . . . . 45 . 48
9.1. Compound Procedure Performance Challenges for Client-Side Caching . . . . . 48
9.2. Proxy Caching . . . . . . . . . . . . . . 45
9.2. Evaluation of a Compound Request . . . . . . . 49
9.3. Delegation and Callbacks . . . . . 45
10. NFS Version 4 Procedures . . . . . . . . . . . 50
9.3.1. Delegation Recovery . . . . . 47
10.1. Procedure 0: NULL - No Operation . . . . . . . . . . . 47
10.2. Procedure 1: COMPOUND - Compound Operations . 51
9.4. Data Caching . . . . . 48
10.3. Procedure 2: ACCESS - Check Access Permission . . . . . 50
10.4. Procedure 3: CLOSE - Close File . . . . . . . . . . . . 53
10.5. Procedure 4: COMMIT - Commit Cached
9.4.1. Data Caching and OPENs . . . . . . . 55
10.6. Procedure 5: CREATE - Create a Non-Regular File Object 58
10.7. Procedure 6: GETATTR - Get Attributes . . . . . . . . . 62
10.8. Procedure 7: GETFH - Get Current Filehandle 53
9.4.2. Data Caching and File Locking . . . . . . 64
10.9. Procedure 8: LINK - Create Link to a . . . . . . 54
9.4.3. Data Caching and Mandatory File Locking . . . . . . . 66
10.10. Procedure 9: LOCK - Create Lock 55
9.4.4. Data Caching and File Identity . . . . . . . . . . . 68
10.11. Procedure 10: LOCKT - Test For Lock . 56
9.5. Open Delegation . . . . . . . . 70
10.12. Procedure 11: LOCKU - Unlock File . . . . . . . . . . 72
10.13. Procedure 12: LOOKUP - Lookup Filename . . 57
9.5.1. Open Delegation and Data Caching . . . . . . 74
10.14. Procedure 13: LOOKUPP - Lookup Parent Directory . . . 77
10.15. Procedure 14: NVERIFY - Verify Difference in Attributes 79
10.16. Procedure 15: OPEN - . . 59
9.5.2. Open a Regular Delegation and File Locks . . . . . . . 81
10.17. Procedure 16: OPENATTR - Open Named Attribute Directory 86
10.18. Procedure 17: PUTFH - Set Current Filehandle . . . . . 88
10.19. Procedure 18: PUTPUBFH - Set Public Filehandle 60
9.5.3. Recall of Open Delegation . . . . 89
10.20. Procedure 19: PUTROOTFH - Set Root Filehandle . . . . 90
10.21. Procedure 20: READ - Read from File . . . . . . 60
9.5.4. Delegation Revocation . . . 91
10.22. Procedure 21: READDIR - Read Directory . . . . . . . . 94
10.23. Procedure 22: READLINK - Read Symbolic Link . . . . . 98
10.24. Procedure 23: REMOVE - Remove Filesystem Object 63
9.6. Data Caching and Revocation . . . 100
10.25. Procedure 24: RENAME - Rename Directory Entry . . . . 102
10.26. Procedure 25: RENEW - Renew a Lease . . . . . . . 63
9.6.1. Revocation Recovery for Write Open Delegation . . 105
10.27. Procedure 25: RESTOREFH - Restore Saved Filehandle . . 106
10.28. Procedure 27: SAVEFH - Save Current Filehandle 63
9.7. Attribute Caching . . . . 108
10.29. Procedure 28: SECINFO - Obtain Available Security . . 109
10.30. Procedure 29: SETATTR - Set Attributes . . . . . . . . 111
10.31. Procedure 30: SETCLIENTID - Negotiated Clientid . . . 114
10.32. Procedure 31: VERIFY - Verify Same Attributes . . 64
9.8. Name Caching . . 116
10.33. Procedure 32: WRITE - Write to File . . . . . . . . . 118
11. Locking notes . . . . . . . . . . . 65
9.9. Directory Caching . . . . . . . . . . . 123
11.1. Short and long leases . . . . . . . . 66
10. Defined Error Numbers . . . . . . . . . 123
11.2. Clocks and leases . . . . . . . . . 68
11. NFS Version 4 Requests . . . . . . . . . . 123
11.3. Locks and lease times . . . . . . . 73
11.1. Compound Procedure . . . . . . . . . . 123
11.4. Locking of directories and other meta-files . . . . . . 124
11.5. Proxy servers and leases . . 73
11.2. Evaluation of a Compound Request . . . . . . . . . . . 73
12. NFS Version 4 Procedures . . 124
11.6. Locking and the new latency . . . . . . . . . . . . . . 124

Expires: December 1999 [Page 4]

Draft Protocol Specification NFS version 4 June 1999

12. Internationalization 75
12.1. Procedure 0: NULL - No Operation . . . . . . . . . . . 75
12.2. Procedure 1: COMPOUND - Compound Operations . . . . . . 76
12.2.1. Operation 2: ACCESS - Check Access Rights . 125
12.1. Universal Versus Local Character Sets . . . . . 78
12.2.2. Operation 3: CLOSE - Close File . . . . 125
12.2. Overview of Universal Character Set Standards . . . . . 126
12.3. Difficulties with UCS-4, UCS-2, Unicode . . 82
12.2.3. Operation 4: COMMIT - Commit Cached Data . . . . . . 127
12.4. UTF-8 and its solutions 84
12.2.4. Operation 5: CREATE - Create a Non-Regular File Object 87
12.2.5. Operation 6: DELEGPURGE - Purge Delegations Awaiting
Recovery . . . . . . . . . . . . . . . . 128
13. Security Considerations . . . . . . 91
12.2.6. Operation 7: DELEGRETURN - Return Delegation . . . . 92
12.2.7. Operation 8: GETATTR - Get Attributes . . . . . . . 129
14. . 93

Expires: March 2000 [Page 4]

Draft Protocol Specification NFS Version version 4 RPC definition file September 1999

12.2.8. Operation 9: GETFH - Get Current Filehandle . . . . . 95
12.2.9. Operation 10: LINK - Create Link to a File . . . . . 97
12.2.10. Operation 11: LOCK - Create Lock . . 130
15. Bibliography . . . . . . . . 99
12.2.11. Operation 12: LOCKT - Test For Lock . . . . . . . . 101
12.2.12. Operation 13: LOCKU - Unlock File . . . . . . 151
16. Authors and Contributors . . . 103
12.2.13. Operation 14: LOOKUP - Lookup Filename . . . . . . . 105
12.2.14. Operation 15: LOOKUPP - Lookup Parent Directory . . 108
12.2.15. Operation 16: NVERIFY - Verify Difference in
Attributes . . . . 155
16.1. Contributors . . . . . . . . . . . . . . . . . 110
12.2.16. Operation 17: OPEN - Open a Regular File . . . . 155
16.2. Editor's Address . . 112
12.2.17. Operation 18: OPENATTR - Open Named Attribute
Directory . . . . . . . . . . . . . . . . . . . 155
16.3. Authors' Addresses . . 120
12.2.18. Operation 19: PUTFH - Set Current Filehandle . . . . 122
12.2.19. Operation 20: PUTPUBFH - Set Public Filehandle . . . 124
12.2.20. Operation 21: PUTROOTFH - Set Root Filehandle . . . 125
12.2.21. Operation 22: READ - Read from File . . . . . . . . 155
17. Full Copyright Statement 126
12.2.22. Operation 23: READDIR - Read Directory . . . . . . . 129
12.2.23. Operation 24: READLINK - Read Symbolic Link . . . . 133
12.2.24. Operation 25: REMOVE - Remove Filesystem Object . . 135
12.2.25. Operation 26: RENAME - Rename Directory Entry . . . 157

Expires: December 1999 [Page 5]

Draft Protocol Specification NFS version 4 June 1999

1. Introduction

NFS version 4 is 137
12.2.26. Operation 27: RENEW - Renew a further revision of the NFS protocol defined
already by versions 2 [RFC1094] and 3 [RFC1813]. It retains the Lease . . . . . . . . 140
12.2.27. Operation 28: RESTOREFH - Restore Saved Filehandle . 141
12.2.28. Operation 29: SAVEFH - Save Current Filehandle . . . 143
12.2.29. Operation 30: SECINFO - Obtain Available Security . 145
12.2.30. Operation 31: SETATTR - Set Attributes . . . . . . . 147
12.2.31. Operation 32: SETCLIENTID - Negotiated Clientid . . 150
12.2.32. Operation 33: VERIFY - Verify Same Attributes . . . 152
12.2.33. Operation 34: WRITE - Write to File . . . . . . . . 154
13. NFS Version 4 Callback Procedures . . . . . . . . . . . . 159
13.1. Procedure 0: CB_NULL - No Operation . . . . . . . . . . 159
13.2. Procedure 1: CB_COMPOUND - Compound Operations . . . . 160
13.2.1. Procedure 2: CB_GETATTR - Get Attributes . . . . . . 162
13.2.2. Procedure 3: CB_RECALL - Recall an Open Delegation . 164
14. Locking notes . . . . . . . . . . . . . . . . . . . . . . 166
14.1. Short and long leases . . . . . . . . . . . . . . . . . 166
14.2. Clocks and leases . . . . . . . . . . . . . . . . . . . 166
14.3. Locks and lease times . . . . . . . . . . . . . . . . . 166
14.4. Locking of directories and other meta-files . . . . . . 167
14.5. Proxy servers and leases . . . . . . . . . . . . . . . 167
14.6. Locking and the new latency . . . . . . . . . . . . . . 167
15. Internationalization . . . . . . . . . . . . . . . . . . 168
15.1. Universal Versus Local Character Sets . . . . . . . . . 168
15.2. Overview of Universal Character Set Standards . . . . . 169
15.3. Difficulties with UCS-4, UCS-2, Unicode . . . . . . . . 170
15.4. UTF-8 and its solutions . . . . . . . . . . . . . . . . 171
16. Security Considerations . . . . . . . . . . . . . . . . . 172
17. NFS Version 4 RPC definition file . . . . . . . . . . . . 173
18. Bibliography . . . . . . . . . . . . . . . . . . . . . . 200

Expires: March 2000 [Page 5]

Draft Protocol Specification NFS version 4 September 1999

19. Authors and Contributors . . . . . . . . . . . . . . . . 204
19.1. Contributors . . . . . . . . . . . . . . . . . . . . . 204
19.2. Editor's Address . . . . . . . . . . . . . . . . . . . 204
19.3. Authors' Addresses . . . . . . . . . . . . . . . . . . 204
20. Full Copyright Statement . . . . . . . . . . . . . . . . 206

Expires: March 2000 [Page 6]

Draft Protocol Specification NFS version 4 September 1999

1. Introduction

NFS version 4 is a further revision of the NFS protocol defined
already by versions 2 [RFC1094] and 3 [RFC1813]. It retains the
essential characteristics of previous versions: stateless design versions: design for easy
recovery, independent of transport protocols, operating systems and
filesystems, simplicity, and good performance. The NFS version 4
revision has the following goals:

o Improved access and good performance on the Internet.

The protocol is designed to transit firewalls easily, perform
well where latency is high and bandwidth is low, and scale to
very large numbers of clients per server.

o Strong security with negotiation built into the protocol.

The protocol builds on the work of the ONCRPC working group in
supporting the RPCSEC_GSS protocol. Additionally NFS version 4
provides a mechanism to allow clients and servers to negotiate
security and require clients and servers to support a minimal
set of security schemes.

o Good cross-platform interoperability.

The protocol features a filesystem model that provides a useful,
common set of features that does not unduly favor one filesystem
or operating system over another.

o Designed for protocol extensions.

The protocol is designed to accept standard extensions that do
not compromise backward compatibility.

Expires: March 2000 [Page 7]

Draft Protocol Specification NFS version 4 September 1999

2. RPC and Security Flavor

The NFS version 4 protocol is a Remote Procedure Call (RPC)
application that uses RPC version 2 and the corresponding eXternal
Data Representation (XDR) as defined in [RFC1831] and [RFC1832]. The
RPCSEC_GSS security flavor as defined in [RFC2203] MUST be used as
the mechanism to deliver stronger security to NFS version 4.

2.1. Ports and Transports

Historically, NFS version 2 and version 3 servers have resided on
UDP/TCP port 2049. Port 2049 is a IANA registered port number for NFS
and therefore will continue to be used for NFS version 4. Using the
well known port for NFS services means the NFS client will not need
to use the RPC binding protocols as described in [RFC1833]; this will
allow NFS to transit firewalls.

The NFS server SHOULD offer its RPC service via TCP as the primary
transport. The server SHOULD also provide UDP for RPC service. The
NFS client SHOULD also have a preference for TCP usage but may supply
a mechanism to override TCP in favor of UDP as the RPC transport.

2.2. Security Flavors

Traditional RPC implementations have included AUTH_NONE, AUTH_SYS,
AUTH_DH, and AUTH_KRB4 as security flavors. With [RFC2203] an
additional security flavor of RPCSEC_GSS has been introduced which
uses the functionality of GSS-API [RFC2078]. This allows for the use
of varying security mechanisms by the RPC layer without the
additional implementation overhead of adding RPC security flavors.
For NFS version 4, the RPCSEC_GSS security flavor MUST be used to
enable the mandatory security mechanism. The flavors AUTH_NONE,
AUTH_SYS, and AUTH_DH MAY be implemented as well.

2.2.1. Security mechanisms for NFS version 4

The use of RPCSEC_GSS requires selection of: mechanism, quality of
protection, and service (authentication, integrity, privacy). The
remainder of this document will refer to these three parameters of
the RPCSEC_GSS security as the security triple.

2.2.1.1. Kerberos V5 as security triple

The Kerberos V5 GSS-API mechanism as described in [RFC1964] MUST be
implemented and provide the following security triples.

columns:

Expires: March 2000 [Page 8]

Draft Protocol Specification NFS version 4 September 1999

1 == number of pseudo flavor
2 == name of pseudo flavor
3 == mechanism's OID
4 == mechanism's algorithm(s)
5 == RPCSEC_GSS service

1 2 3 4 5
-----------------------------------------------------------------------
390003 krb5 1.2.840.113554.1.2.2 DES MAC MD5 rpc_gss_svc_none
390004 krb5i 1.2.840.113554.1.2.2 DES MAC MD5 rpc_gss_svc_integrity
390005 krb5p 1.2.840.113554.1.2.2 DES MAC MD5 rpc_gss_svc_privacy
for integrity,
and 56 bit DES
for privacy.

Note that the pseudo flavor is presented here as a mapping aid to the
implementor. Because this NFS protocol includes a method to
negotiate security and it understands the GSS-API mechanism, the
pseudo flavor is not needed. The pseudo flavor is needed for NFS
version 3 since the security negotiation is done via the MOUNT
protocol.

For a discussion of NFS' use of RPCSEC_GSS and Kerberos V5, please
see [RFC2623].

2.2.1.2. <another security triple>

Another GSS-API mechanism will need to be specified here
along with the corresponding security triple(s).

2.3. Security Negotiation

With the NFS version 4 server potentially offering multiple security
mechanisms, the client will need a way to determine or negotiate
which mechanism is to be used for its communication with the server.
The NFS server may have multiple points within its file system name
space that are available for use by NFS clients. In turn the NFS
server may be configured such that each of these entry points may
have different or multiple security mechanisms in use.

The security negotiation between client and server must be done with
a secure channel to eliminate the possibility of a third party
intercepting the negotiation sequence and forcing the client and
server to choose a lower level of security than required/desired.

Expires: March 2000 [Page 9]

Draft Protocol Specification NFS version 4 September 1999

2.3.1. Security Error

Based on the assumption that each NFS version 4 client and server
must support a minimum set of security (i.e. Kerberos-V5 under
RPCSEC_GSS, <ed: add other>), the NFS client will start its
communication with the server with one of the minimal security
triples. During communication with the server, the client may
receive an NFS error of NFS4ERR_WRONGSEC. This error allows the
server to notify the client that the security triple currently being
used is not appropriate for access to the server's file system
resources. The client is then responsible for determining what
security triples are available at the server and choose one which is
appropriate for the client.

2.3.2. SECINFO

The new procedure SECINFO (see SECINFO procedure definition) will
allow the client to determine, on a per filehandle basis, what
security triple is to be used for server access. In general, the
client will not have to use the SECINFO procedure except during
initial communication with the server or when the client crosses
policy boundaries at the server. It could happen that the server's
policies change during the client's interaction therefore forcing the
client to negotiate a new security triple.

Expires: March 2000 [Page 10]

Draft Protocol Specification NFS version 4 September 1999

3. File handles

The file handle in the NFS protocol is a per server unique identifier
for a file system object. The contents of the file handle are opaque
to the client. Therefore, the server is responsible for translating
the file handle to an internal representation of the file system
object. Since the file handle is the client's reference to an object
and the client may cache this reference, the server should not reuse
a file handle for another file system object. If the server needs to
reuse a file handle value, the time elapsed before reuse SHOULD be
large enough that it is likely the client no longer has a cached copy
of the reused file handle value.

3.1. Obtaining the First File Handle

The procedures of the NFS protocol are defined in terms of one or
more file handles. Therefore, the client needs a file handle to
initiate communication with the server. With NFS version 2 [RFC1094]
and NFS version 3 [RFC1813], there exists an ancillary protocol to
obtain this first file handle. The MOUNT protocol, RPC program
number 100005, provides the mechanism of translating a string based
file system path name to a file handle which can then be used by the
NFS protocols.

The MOUNT protocol has deficiencies in the area of security and use
via firewalls. This is one reason that the use of the public file
handle was introduced [RFC2054] [RFC2055]. With the use of public
file handle in combination with the LOOKUP procedure in NFS version 2
and 3, it has been demonstrated that the MOUNT protocol is
unnecessary for viable interaction between NFS client and server.

Therefore, NFS version 4 will not use an ancillary protocol for
translation from string based path names to a file handle. Two
special file handles will be used as starting points for the NFS
client.

3.1.1. Root File Handle

The first of the special file handles is the ROOT file handle. The
ROOT file handle is the "conceptual" root of the file system name
space at the NFS server. The client uses or starts with the ROOT
file handle by employing the PUTROOTFH procedure. The PUTROOTFH
procedure instructs the server to set the "current" file handle to
the ROOT of the server's file tree. Once this PUTROOTFH procedure is
used, the client can then traverse the entirety of the server's file
tree with the LOOKUP procedure. A complete discussion of the server
name space is in section 7, "NFS Server Name Space".

Expires: March 2000 [Page 11]

Draft Protocol Specification NFS version 4 September 1999

3.1.2. Public File Handle

The second special file handle is the PUBLIC file handle. Unlike the
ROOT file handle, the PUBLIC file handle may be bound or represent an
arbitrary file system object at the server. The server is
responsible for this binding. It may be that the PUBLIC file handle
and the ROOT file handle refer to the same file system object.
However, it is up to the administrative software at the server and
the policies of the server administrator to define the binding of the
PUBLIC file handle and server file system object. The client may not
make any assumptions about this binding.

3.2. File Handle Types

In NFS version 2 and 3, there was one type of file handle with a
single set of semantics. NFS version 4 introduces a new type of file
handle in an attempt to accommodate certain server environments. The
first type of file handle is 'persistent'. The semantics of a
persistent file handle are the same as the file handles of NFS
version 2 and 3. The second or new type of file handle is the
'volatile' file handle.

The volatile file handle type is being introduced to address server
functionality or implementation issues which prevent correct or
feasible implementation of a persistent file handle. Some server
environments do not provide a file system level invariant that can be
used to construct a persistent file handle. The underlying server
file system may not provide the invariant or the server's file system
APIs may not provide access to the needed invariant. Volatile file
handles may ease the implementation of server functionality such as
hierarchical storage management or file system reorganization or
migration. However, the volatile file handle increases the
implementation burden for the client but this increased burden is
deemed acceptable based on the overall gains achieved by the
protocol.

Since the client will have different paths of logic to handle
persistent and volatile file handles, a file attribute is defined
which may be used by the client to determine the file handle types
being returned by the server.

3.2.1. General Properties of a File Handle

The file handle contains all the information the server needs to
distinguish an individual file. To the client, the file handle is
opaque. The client stores file handles for use in a later request and

Expires: March 2000 [Page 12]

Draft Protocol Specification NFS version 4 September 1999

can compare two file handles from the same server for equality by
doing a byte-by-byte comparison, but MUST NOT otherwise interpret the
contents of file handles. If two file handles from the same server
are equal, they MUST refer to the same file, but if they are not
equal, no conclusions can be drawn. Servers SHOULD try to maintain a
one-to-one correspondence between file handles and files but this is
not required. Clients MUST only use file handle comparisons only to
improve performance, not for correct behavior.

As an example, in the case that two different path names when
traversed at the server terminate at the same file system object, the
server SHOULD return the same file handle for each path. This can
occur if a hard link is used to create two file names which refer to
the same underlying file object and associated data. For example, if
paths /a/b/c and /a/d/c refer to the same file, the server SHOULD
return the same file handle for both path names traversals.

3.2.2. Persistent File Handle

A persistent file handle is defined as having a persistent value for
the lifetime of the file system object to which it refers. Once the
server creates the file handle for a file system object, the server
MUST return the same file handle for the object for the lifetime of
the object. If the server restarts or reboots, or the filesystem is
migrated, the NFS server must honor and present the same file handle
value as it did in the server's previous instantiation.

The persistent file handle will be become stale or invalid when the
file system object is removed. When the server is presented with a
persistent file handle that refers to a deleted object, it MUST
return an error of NFS4ERR_STALE. A file handle may become stale
when the file system containing the object is no longer available.
The file system may become unavailable if it exists on removable
media and the media is no longer available at the server or the file
system in whole has been destroyed or the file system has simply been
removed from the server's name space (i.e. unmounted in a Unix
environment).

3.2.3. Volatile File Handle

A volatile file handle does not share the same longevity attributes
of the persistent file handle. The server may determine that a
volatile file handle is no longer valid at many different points in
time. If the server can definitively determine that a volatile file
handle refers to an object that has been removed, the server should
return NFS4ERR_STALE to the client (as is the case for persistent

Expires: March 2000 [Page 13]

Draft Protocol Specification NFS version 4 September 1999

file handles). In all other cases where the server determines that a
volatile file handle can no longer be used, it should return an error
of NFS4ERR_EXPIRED.

The following table shows the most common points at which a volatile
file handle may expire. This table represents the view from the
client's perspective and as such provides columns for when the file
may be open or closed by the client.

Server Provides Persistent or Volatile File Handle
File Open File Closed
___________________________________________________________________
Restart of Server (note 4) P / V P / V
Filesystem Migration (note 5) P / V P / V
SHARE/LOCK recovery P / V N/A (note 1)
Client RENAMEs object P / V P / V
Client RENAMEs path to object P / V P / V
Other client RENAMEs object P / V P / V
Other client RENAMEs path to object P / V P / V
Client REMOVEs object P / V (note 2) P / V
Other client REMOVEs object P / V N/A (note 3)

Note 1
If the file is not open, persistence of the file handle is not
applicable for the recovery of SHARE/LOCK.

Note 2
With NFS version 2 and 3, when the client removes a file it has
open it follows the convention of RENAMEing the file to '.nfsXXXX'
until the file is closed. At this point the REMOVE is done at the
server.

If this same model is used for v4 then this entry will be
'N/A'.

Note 3
If the file is not open by the client, then it should not expect
any cached file handle to be valid.

Note 4
The restart of the NFS server signifies when the operating system
or NFS software is (re)started. This also includes High
Availability configurations where a separate operating system
instantiation acquires ownership of the file system resources and
network resources (i.e. disks and IP addresses).

Expires: March 2000 [Page 14]

Draft Protocol Specification NFS version 4 September 1999

Note 5
Filesystem migration may occur in response to an unresponsive
server or when the current server indicates that a filesystem has
moved by returning NFS4ERR_MOVED. In both cases, the attribute
fs_locations designates the new server location for the filesystem.

3.2.4. One Method of Constructing a Volatile File Handle

As mentioned, in some instances a file handle is stale (no longer
valid, perhaps because the file was removed from the server), or it
is expired (the underlying file is valid, but since the file handle
is volatile, it may have expired). Thus the server needs to be able
to return NFS4ERR_STALE in the former case, and NFS4ERR_FHEXPIRED in
the latter case. This can be done by careful construction of the
volatile file handle. One possible implementation follows.

A volatile file handle, while opaque to the client could contain:

[volatile bit = 1 | server boot time | slot | generation number]

o slot is an index in the server volatile file handle table

o generation number is the generation number for the table
entry/slot

If the server boot time is less than the current server boot time,
return NFS4ERR_FHEXPIRED. If slot is out of range, return
NFS4ERR_BADHANDLE. If the generation number does not match, return
NFS4ERR_BADHANDLE.

When the server reboots, the table is gone (it is volatile).

If volatile bit is 0, then it is a persistent file handle with a
different structure following it.

3.3. Client Recovery from File Handle Expiration

With the introduction of the volatile file handle, the client must
take on additional responsibility so that it may prepare itself to
recover from the expiration of a volatile file handle. If the server
returns persistent file handles, the client does not need these
additional steps.

For volatile file handles, most commonly the client will need to
store the component names leading up to and including the file system

Expires: March 2000 [Page 15]

Draft Protocol Specification NFS version 4 September 1999

object in question. With these names, the client should be able to
recover by finding a file handle in the name space that is still
available or by starting at the root of the server's file system name
space.

If the expired file handle refers to an object that has been removed
from the file system, obviously the client will not be able to
recover from the expired file handle.

It is also possible that the expired file handle refers to a file
that has been renamed. If the file was renamed by another client,
again it is possible that the original client will not be able to
recover. However, in the case that the client itself is renaming the
file and the file is open, it is possible that the client may be able
to recover. The client can determine the new path name based on the
processing of the rename request. The client can then regenerate the
new file handle based on the new path name. The client could also
use the compound operation mechanism to construct a set of operations
like:
RENAME A B
LOOKUP B
GETFH

Expires: March 2000 [Page 16]

Draft Protocol Specification NFS version 4 September 1999

4. Basic Data Types

Arguments and results from operations will be described in terms of
basic XDR types defined in [RFC1832]. The following data types will
be defined in terms of basic XDR types:

filehandle: opaque <128>

An NFS version 4 filehandle. A filehandle with zero length is
recognized as a "public" filehandle.

utf8string: opaque <>

A counted array of octets that contains a UTF-8 string.

Note: Section 11, Internationalization, covers the rational of
using UTF-8.

bitmap: uint32 <>

A counted array of 32 bit integers used to contain bit values.
The position of the integer in the array that contains bit n can
be computed from the expression (n / 32) and its bit within that
integer is (n mod 32).

0 1
+-----------+-----------+-----------+--
| count | 31 .. 0 | 63 .. 32 |
+-----------+-----------+-----------+--

createverf: opaque<8>

Verify used for exclusive create semantics

nfstime4
struct nfstime4 {
int64_t seconds;
uint32_t nseconds;
}

The nfstime4 structure gives the number of seconds and
nanoseconds since midnight or 0 hour January 1, 1970 Coordinated
Universal Time (UTC). Values greater than zero for the seconds
field denote dates after the 0 hour January 1, 1970. Values
less than zero for the seconds field denote dates before the 0
hour January 1, 1970. In both cases, the nseconds field is to
be added to the seconds field for the final time representation.
For example, if the time to be represented is one-half second

Expires: March 2000 [Page 17]

Draft Protocol Specification NFS version 4 September 1999

before 0 hour January 1, 1970, the seconds field would have a
value of negative one (-1) and the nseconds fields would have a
value of one-half second (500000000). Values greater than
999,999,999 for nseconds are considered invalid.

This data type is used to pass time and date information. A
server converts to and from local time when processing time
values, preserving as much accuracy as possible. If the
precision of timestamps stored for a file system object is less
than defined, loss of precision can occur. An adjunct time
maintenance protocol is recommended to reduce client and server
time skew.

specdata4
struct specdata4 {
uint32_t specdata1;
uint32_t specdata2;
}

This data type represents additional information for the device
file types NFCHR and NFBLK.

Expires: March 2000 [Page 18]

Draft Protocol Specification NFS version 4 September 1999

5. File Attributes

To meet the NFS Version 4 requirements of extensibility and increased
interoperability with non-Unix platforms, attributes must be handled
in a more flexible manner. The NFS Version 3 fattr3 structure
contained a fixed list of attributes that not all clients and servers
are able to support or care about, which cannot be extended as new
needs arise, and which provides no way to indicate non-support. With
NFS Version 4, the client will be able to ask what attributes the
server supports, and will be able to request only those attributes in
which it is interested.

To this end, attributes will be divided into three groups: mandatory,
recommended and named. Both mandatory and recommended attributes are
supported in the NFS V4 protocol by a specific and well-defined
encoding, and are identified by number. They are requested by
setting a bit in the bit vector sent in the GETATTR request; the
server response includes a bit vector to list what attributes were
returned in response. New mandatory or recommended attributes may be
added to the NFS protocol between revisions by publishing a
standards-track RFC which allocates a new attribute number value and
defines the encoding for the attribute.

Named attributes are accessed by the new OPENATTR operation, which
accesses a hidden directory of attributes associated with a
filesystem object. OPENATTR takes a filehandle for the object and
returns the filehandle for the attribute hierarchy, which is a
directory object accessible by LOOKUP or READDIR, and which contains
files whose names represent the named attributes and whose data bytes
are the value of the attribute. For example:

LOOKUP "foo" ; look up file
GETATTR attrbits
OPENATTR ; access foo's named attributes
LOOKUP "x11icon" ; look up specific attribute
READ 0,4096 ; read stream of bytes

Named attributes are intended primarily for data needed by
applications rather than by an NFS client implementation per se; NFS
implementors are strongly encouraged to define their new attributes
as recommended attributes by bringing them to the working group.

The set of attributes which are classified as mandatory is
deliberately small, since servers must do whatever it takes to
support them. The recommended attributes may be unsupported, though
a server should support as many as it can. Attributes are deemed

Expires: March 2000 [Page 19]

Draft Protocol Specification NFS version 4 September 1999

mandatory if the data is both needed by a large number of clients and
is not otherwise reasonably computable by the client when support is
not provided on the server.

5.1. Mandatory Attributes

These MUST be supported by every NFS Version 4 client and server in
order to ensure a minimum level of interoperability. The server must
store and return these attributes, and the client must be able to
function with an attribute set limited to these attributes, though
some operations may be impaired or limited in some ways in this case.
A client may ask for any of these attributes to be returned by
setting a bit in the GETATTR request, and the server must return
their value.

5.2. Recommended Attributes

These attributes are understood well enough to warrant support in the
NFS Version 4 protocol, though they may not be supported on all
clients and servers. A client may ask for any of these attributes to
be returned by setting a bit in the GETATTR request, but must be able
to deal with not receiving them. A client may ask for the set of
attributes the server supports and should not request attributes the
server does not support. A server should be tolerant of requests for
unsupported attributes, and simply not return them, rather than
considering the request an error. It is expected that servers will
support all attributes they comfortably can, and only fail to support
attributes which are difficult to support in their operating
environments. A server should provide attributes whenever they don't
have to "tell lies" to the client. For example, a file modification
time should be either an accurate time or should not be supported by
the server. This will not always be comfortable to clients but it
seems that the client has a better ability to fabricate or construct
an attribute or do without.

Most attributes from NFS V3's FSINFO, FSSTAT and PATHCONF procedures
have been added as recommended attributes, so that filesystem info
may be collected via the filehandle of any object the filesystem.
This renders those procedures unnecessary in NFS V4.

5.3. Named Attributes

These attributes are not supported by direct encoding in the NFS
Version 4 protocol but are accessed by string names rather than
numbers and correspond to an uninterpreted stream of bytes which are

Expires: March 2000 [Page 20]

Draft Protocol Specification NFS version 4 September 1999

stored with the filesystem object. The namespace for these
attributes may be accessed by using the OPENATTR operation to get a
filehandle for a virtual "attribute directory" and using READDIR and
LOOKUP operations on this filehandle. Named attributes may then be
examined or changed by normal READ and WRITE and CREATE operations on
the filehandles returned from READDIR and LOOKUP. Named attributes
may have attributes, for example, a security label may have access
control information in its own right.

It is recommended that servers support arbitrary named attributes. A
client should not depend on the ability to store any named attributes
in the server's filesystem. If a server does support named
attributes, a client which is also able to handle them should be able
to copy a file's data and meta-data with complete transparency from
one location to another; this would imply that there should be no
attribute names which will be considered illegal by the server.

Names of attributes will not be controlled by a standards body.
However, vendors and application writers are encouraged to register
attribute names and the interpretation and semantics of the stream of
bytes via informational RFC so that vendors may interoperate where
common interests exist.

Expires: March 2000 [Page 21]

Draft Protocol Specification NFS version 4 September 1999

5.4. Mandatory Attributes - Definitions

Name # DataType Access Description
___________________________________________________________________
supp_attr 0 bitmap READ
The bit vector which
would retrieve all
mandatory and
recommended attributes
which may be requested
for this object.

The client must ask
this question to
request correct
attributes.

object_type 1 nfs4_ftype READ
The type of the object
(file, directory,
symlink)

The client cannot
handle object
correctly without
type.

persistent_fh 2 boolean READ
Is the filehandle for
this object
persistent?

Server should know if
the filehandles being
provided are
persistent or not. If
the server is not able
to make this
determination, then it
can choose volatile or
non-persistent.

Expires: March 2000 [Page 22]

Draft Protocol Specification NFS version 4 September 1999

change 3 uint64 READ
A value created by the
server that the client
can use to determine
if a file data,
directory contents or
attributes have been
modified. The server
can just return the
file mtime in this
field though if a more
precise value exists
then it can be
substituted, for
instance, a sequence
number.

Necessary for any
useful caching, likely
to be available.

object_size 4 uint64 R/W
The size of the object
in bytes.

Could be very
expensive to derive,
likely to be
available.

link_support 5 boolean READ
Does the object's
filesystem supports
hard links?

Server can easily
determine if links are
supported.

symlink_support 6 boolean READ
Does the object's
filesystem supports
symbolic links?

Server can easily
determine if links are
supported.

named_attr 7 boolean READ
Does this object have
named attributes?

Expires: March 2000 [Page 23]

Draft Protocol Specification NFS version 4 September 1999

fsid 8 fsid4 READ
Unique filesystem
identifier for the
filesystem holding
this object. fsid
contains major and
minor components each
of which are uint64.

unique_handles 9 boolean READ
Are two distinct
filehandles guaranteed
to refer to two
different file system
objects?

lease_time 10 uint32 READ
Duration of leases at
server in seconds.

Expires: March 2000 [Page 24]

Draft Protocol Specification NFS version 4 September 1999

5.5. Recommended Attributes - Definitions

Name # Data Type Access Description
_____________________________________________________________________
ACL 11 nfsacl4 R/W
The access control
list for the object.
[The nature and
format of ACLs is
still to be
determined.]

archive 12 boolean R/W
Whether or not this
file has been
archived since the
time of last
modification
(deprecated in favor
of backup_time).

cansettime 13 boolean READ
Whether or not this
object's filesystem
can fill in the times
on a SETATTR request
without an explicit
time.

case_insensitive 14 boolean READ
Are filename
comparisons on this
filesystem case
insensitive?

case_preserving 15 boolean READ
Is filename case on
this filesystem
preserved?

chown_restricted 16 boolean READ
Will a request to
change ownership be
honored?

filehandle 17 nfs4_fh READ
The filehandle of
this object
(primarily for
readdir requests).

Expires: March 2000 [Page 25]

Draft Protocol Specification NFS version 4 September 1999

fileid 18 uint64 READ
A number uniquely
identifying the file
within the
filesystem.

files_avail 19 uint64 READ
File slots available
to this user on the
filesystem containing
this object - this
should be the
smallest relevant
limit.

files_free 20 uint64 READ
Free file slots on
the filesystem
containing this
object - this should
be the smallest
relevant limit.

files_total 21 uint64 READ
Total file slots on
the filesystem
containing this
object.

fs_locations 22 fs_locations READ
Locations where this
filesystem may be
found. If the server
returns NFS4ERR_MOVED
as an error, this
attribute must be
supported.

hidden 23 boolean R/W
Is file considered
hidden?

homogeneous 24 boolean READ
Whether or not this
object's filesystem
is homogeneous, i.e.
whether pathconf is
the same for all
filesystem objects.

maxfilesize 25 uint64 READ
Maximum supported
file size for the
filesystem of this
object.

Expires: March 2000 [Page 26]

Draft Protocol Specification NFS version 4 September 1999

maxlink 26 uint32 READ
Maximum number of
links for this
object.

maxname 27 uint32 READ
Maximum filename size
supported for this
object.

maxread 28 uint64 READ
Maximum read size
supported for this
object.

maxwrite 29 uint64 READ
Maximum write size
supported for this
object. This
attribute SHOULD be
supported if the file
is writable. Lack of
this attribute can
lead to the client
either wasting
bandwidth or not
receiving the best
performance.

mime_type 30 utf8<> R/W
MIME body
type/subtype of this
object.

mode 31 uint32 R/W
Unix-style permission
bits for this object
(deprecated in favor
of ACLs)

no_trunc 32 boolean READ
If a name longer than
name_max is used,
will an error be
returned or will the
name be truncated?

numlinks 33 uint32 READ
Number of links to
this object.

owner 34 utf8<> R/W
The string name of
the owner of this
object.

Expires: March 2000 [Page 27]

Draft Protocol Specification NFS version 4 September 1999

owner_group 35 utf8<> R/W
The string name of
the group of the
owner of this object.

quota_hard 36 uint64 READ
Number of bytes of
disk space beyond
which the server will
decline to allocate
new space.

quota_soft 37 uint64 READ
Number of bytes of
disk space at which
the client may choose
to warn the user
about limited space.

quota_used 38 uint64 READ
Number of bytes of
disk space occupied
by the owner of this
object on this
filesystem.

rawdev 39 specdata4 READ
Raw device
identifier.

space_avail 40 uint64 READ
Disk space in bytes
available to this
user on the
filesystem containing
this object - this
should be the
smallest relevant
limit.

space_free 41 uint64 READ
Free disk space in
bytes on the
filesystem containing
this object - this
should be the
smallest relevant
limit.

space_total 42 uint64 READ
Total disk space in
bytes on the
filesystem containing
this object.

Expires: March 2000 [Page 28]

Draft Protocol Specification NFS version 4 September 1999

space_used 43 uint64 READ
Number of filesystem
bytes allocated to
this object.

system 44 boolean R/W
Whether or not this
file is a system
file.

time_access 45 nfstime4 R/W
The time of last
access to the object.

time_backup 46 nfstime4 R/W
The time of last
backup of the object.

time_create 47 nfstime4 R/W
The time of creation
of the object. This
attribute does not
have any relation to
the traditional Unix
file attribute
time'.

time_delta 48 nfstime4 READ
Smallest useful
server time
granularity.

time_metadata 49 nfstime4 R/W
The time of last
meta-data
modification of the
object.

time_modify 50 nfstime4 R/W
The time since the
epoch of last
modification to the
object.

version 51 utf8<> R/W
Version number of
this document.

volatility 52 nfstime4 READ
Approximate time
until next expected
change on this
filesystem, as a
measure of
volatility.

Expires: March 2000 [Page 29]

Draft Protocol Specification NFS version 4 September 1999

5.6. Interpreting owner and owner_group

The recommended attributes "owner" and "owner_group" are represented
in terms of a UTF-8 string. To avoid a representation that is tied
to a particular underlying implementation at the client or server,
the use of the UTF-8 string has been chosen. Note that section 6.1
of [RFC2624] provides additional rationale. It is expected that the
client and server will have their own local representation of owner
and owner_group that is used for local storage or presentation to the
end user. Therefore, it is expected that the when these attributes
are transferred between the client and server that the local
representation is translated to a syntax of the form
"user@dns_domain". This will allow for a client and server that do
not use the same local representation the ability to translate to a
common syntax that can be interpreted by both.

The translation is not specified as part of the protocol. This
allows various solutions to be employed. For example, a local
translation table may be consulted that maps between a numeric id to
the user@dns_domain syntax. A name service may also be used to
accomplish the translation. The 'dns_domain' portion of the owner
string is meant to be a DNS domain name. For example, user@ietf.org.

In the case where there is no translation available to the client or
server, the attribute value must be constructed without the '@'.
Therefore, the absence of the @ from the owner or owner_group
attribute signifies that no translation was available and the
receiver of the attribute should not place any special meaning with
the attribute value. Even though the attribute value can not be
translated, it may still be useful. In the case of a client, the
attribute string may be used for local display of ownership.

Expires: March 2000 [Page 30]

Draft Protocol Specification NFS version 4 September 1999

6. Filesystem Migration and Replication

With the use of the recommended attribute "fs_locations", the NFS
version 4 server has a method of providing filesystem migration or
replication services. For the purposes of migration and replication,
a filesystem will be defined as all files that share a given fsid
(major and minor values are the same).

The fs_locations attribute provides a list of filesystem locations.
These locations are specified by providing the server name (either
DNS domain or IP address) and the path name representing the root of
the filesystem. Depending on the type of service being provided, the
list will provide a new or alternate locations for the filesystem.
The client will use this information to redirect its requests to the
new server.

6.1. Replication

It is expected that filesystem replication will be used in the case
of read-only data. Typically, the filesystem will be replicated
amongst two or more servers. The fs_locations attribute will provide
the list of these locations to the client. On first access of the
filesystem, the client should obtain the value of the fs_locations
attribute. If, in the future, the client finds the server
unresponsive, the client may attempt to use another server specified
by fs_locations.

If applicable, the client must take the appropriate steps to recover
valid filehandles from the new server. This is described in more
detail in the following sections.

6.2. Migration

Filesystem migration is used to move a filesystem from one server to
another. Migration is typically used for a filesystem that is
writable and has a single copy. The expected use of migration is for
load balancing or general resource reallocation. The protocol does
not specify how the filesystem will be moved between servers. This
server-to-server transfer mechanism is left to the server
implementor. However, the method used to communicate the migration
event between client and server is specified here.

Once the servers participating in the migration have completed the
move of the filesystem, the error NFS4ERR_MOVED will be returned for
subsequent requests received by the original server. The
NFS4ERR_MOVED error is returned for all operations except GETATTR.

Expires: March 2000 [Page 31]

Draft Protocol Specification NFS version 4 September 1999

Upon receiving the NFS4ERR_MOVED error, the client will obtain the
value of the fs_locations attribute. The client will then use the
contents of the attribute to redirect its requests to the specified
server. To facilitate the use of GETATTR operations such as PUTFH
must also be accepted by the server for the migrated filesystem's
filehandles. Note that if the server returns NFS4ERR_MOVED, the
server MUST support the fs_locations attribute.

If the client requests more attributes than fs_locations, the server
may return fs_locations only. This is to be expected since the
server has migrated the filesystem and may not have a method of
obtaining additional attribute data.

The server implementor needs to be careful in developing a migration
solution. The server must consider all of the state information
clients may have outstanding at the server. This includes but is not
limited to locking/share state, delegation state, and asynchronous
file writes which are represented by WRITE and COMMIT verifiers. The
server should strive to minimize the impact on its clients during and
after the migration process.

6.3. Interpretation of the fs_locations Attribute

The fs_location attribute is structured in the following way:

struct fs_location {
utf8string server<>;
pathname4 rootpath;
};

struct fs_locations {
pathname4 fs_root;
fs_location locations<>;
};

The fs_location struct is used to represent the location of a
filesystem by providing a server name and the path to the root of the
filesystem. For a multi-homed server or a set of servers that use
the same rootpath, an array of server names may be provided. An
entry in the server array is an UTF8 string and represents one of a
traditional DNS host name, IPv4 address, or IPv6 address. It is not
a requirement that all servers that share the same rootpath be listed
in one fs_location struct. The array of server names is provided for
convenience. Servers that share the same rootpath may also be listed
in separate fs_location entries in the fs_locations attribute.

The fs_locations struct and attribute then contains an array of

Expires: March 2000 [Page 32]

Draft Protocol Specification NFS version 4 September 1999

locations. Since the namespace of each server may be constructed
differently, the "fs_root" field is provided. The path represented
by fs_root represents the location of the filesystem in the server's
namespace. Therefore, the fs_root path is only associated with the
server from which the fs_locations attribute was obtained. The
fs_root path is meant to aid the client in locating the filesystem at
the various servers listed.

As an example, there is a replicated file system located at two
servers (servA and servB). At servA the filesystem is located at
path "/a/b/c". At servB the filesystem is located at path "/x/y/z".
In this example the client accesses the filesystem first at servA
with a multi-component lookup path of "/a/b/c/d". Since the client
used a multi-component lookup to obtain the filehandle at "/a/b/c/d",
it is unaware that the filesystem's root is located in servA's
namespace at "/a/b/c". When the client switches to servB, it will
need to determine that the directory it first referenced at servA is
now represented by the path "/x/y/z/d" on servB. To facilitate this,
the fs_locations attribute provided by servA would have a fs_root
value of "/a/b/c" and two entries in fs_location. One entry in
fs_location will be for itself (servA) and the other will be for
servB with a path of "/x/y/z". With this information, the client is
able to substitute "/x/y/z" for the "/a/b/c" at the beginning of its
access path and construct "/x/y/z/d" to use for the new server.

6.4. Filehandle Recovery for Migration or Replication

Filehandles for filesystems that are replicated or migrated have the
same semantics as for filesystems that are not replicated or
migrated. For example, if a filesystem has persistent filehandles
and it is migrated to another server, the filehandle values for the
filesystem will be valid at the new server.

The same is true for a filesystem which is made up of volatile
filehandles. In fact, in this case the client should expect that the
new server will return NFS4ERR_EXPIRED when old filehandles are
presented; the client will need to recover the filehandles
appropriately.

Expires: March 2000 [Page 33]

Draft Protocol Specification NFS version 4 September 1999

7. NFS Server Namespace

7.1. Server Exports

On a UNIX server the name-space describes all the files reachable by
pathnames under the root directory "/". On a Windows NT server the
name-space constitutes all the files on disks named by mapped disk
letters. NFS server administrators rarely make the entire server's
file-system name-space available to NFS clients. Typically, pieces
of the name-space are made available via an "export" feature. In
previous versions of NFS, the root file-handle for each export is
obtained through the MOUNT protocol; the client sends a string that
identifies the export of name-space and the server returns the root
file-handle for it. The MOUNT protocol supports an EXPORTS procedure
that will enumerate the server's exports.

7.2. Browsing Exports

The NFS version 4 protocol provides a root file-handle that clients
can use to obtain file-handles for these exports via a multi-
component LOOKUP. A common user experience is to use a graphical
user interface (perhaps a file "Open" dialog window) to find a file
via progressive browsing through a directory tree. The client must be
able to move from one export to another export via single-component,
progressive LOOKUP operations.

This style of browsing is not well supported by NFS version 2 and 3
protocols. The client expects all LOOKUP operations to remain within
a single server file-system, i.e. the device attribute will not
change. This prevents a client from taking name-space paths that
span exports.

An automounter on the client can obtain a snapshot of the server's
name-space using the EXPORTS procedure of the MOUNT protocol. If it
understands the server's pathname syntax, it can create an image of
the server's name-space on the client. The parts of the name-space
that are not exported by the server are filled in with a "pseudo
file-system" that allows the user to browse from one mounted file-
system to another. There is a drawback to this representation of the
server's name-space on the client: it is static. If the server
administrator adds a new export the client will be unaware of it.

Expires: March 2000 [Page 34]

Draft Protocol Specification NFS version 4 September 1999

7.3. Server Pseudo File-System

NFS version 4 servers avoid this name-space inconsistency by
presenting all the exports within the framework of a single server
name-space. An NFS version 4 client uses LOOKUP and READDIR
operations to browse seamlessly from one export to another. Portions
of the server name-space that are not exported are bridged via a
"pseudo file-system" that provides a view of exported directories
only. A pseudo file-system has a unique fsid and behaves like a
normal, read-only file-system.

Based on the construction of the server's name space, it is
possible that multiple pseudo filesystems may exist. For
example,

/a pseudo filesystem
/a/b real filesystem
/a/b/c pseudo filesystem
/a/b/c/d real filesystem

Need to discuss the ramifications of multiple pseudo
filesystems.

7.4. Multiple Roots

DOS, Windows 95, 98 and NT are sometimes described as having
"multiple roots". File-Systems are commonly represented as disk
letters. MacOS represents file-systems as top-level names. NFS
version 4 servers for these platforms can construct a pseudo file-
system above these root names so that disk letters or volume names
are simply directory names in the pseudo-root.

7.5. Filehandle Volatility

The nature of the server's pseudo file-system is that it is a logical
representation of file-system(s) available from the server.
Therefore, the pseudo file-system is most likely constructed
dynamically when the NFS version 4 is first instantiated. It is
expected the pseudo file-system may not have an on-disk counterpart
from which persistent filehandles could be constructed. Even though
it is preferable that the server provide persistent filehandles for
the pseudo file-system, the NFS client should expect that pseudo
file-system file-handles are volatile. This can be confirmed by
checking the associated "persistent_fh" attribute for those

Expires: March 2000 [Page 35]

Draft Protocol Specification NFS version 4 September 1999

filehandles in question. If the filehandles are volatile, the NFS
client must be prepared to recover a filehandle value (i.e. with a v4
multi-component LOOKUP) when receiving an error of NFS4ERR_FHEXPIRED.

7.6. Exported Root

If the server's root file-system is exported, it might be easy to
conclude that a pseudo-file-system is not needed. This would be
wrong. Assume the following file-systems on a server:

/ disk1 (exported)
/a disk2 (not exported)
/a/b disk3 (exported)

Because disk2 is not exported, disk3 cannot be reached with simple
LOOKUPs. The server must bridge the gap with a pseudo-file-system.

7.7. Mount Point Crossing

The server file-system environment may be constructed in such a way
that one file-system contains a directory which is 'covered' or
mounted upon by a second file-system. For example:

/a/b (file system 1)
/a/b/c/d (file system 2)

The pseudo file-system for this server may be constructed to look
like:

/ (place holder/not exported)
/a/b (file system 1)
/a/b/c/d (file system 2)

It is the server's responsibility to present the pseudo file-system
that is complete to the client. If the client sends a lookup request
for the path "/a/b/c/d", the server's response is the filehandle of
the file system "/a/b/c/d". In previous versions of NFS, the server
would respond with the directory "/a/b/d/d" within the file-system
"/a/b".

The NFS client will be able to determine if it crosses a server mount
point by a change in the value of the "fsid" attribute.

Expires: March 2000 [Page 36]

Draft Protocol Specification NFS version 4 September 1999

7.8. Security Policy and Namespace Presentation

The application of the server's security policy needs to be carefully
considered by the implementor. One may choose to limit the
viewability of portions of the pseudo file-system based on the
server's perception of the client's ability to authenticate itself
properly. However with the support of multiple security mechanisms
and the ability to negotiate the appropriate use of these mechanisms,
the server is unable to properly determine if a client will be able
to authenticate itself. If, based on its policies, the server
chooses to limit the contents of the pseudo file-system, the server
may effectively hide file-systems from a client that may otherwise
have legitimate access.

7.9. Summary

NFS version 4 provides LOOKUP and READDIR operations for
easy recovery, independent browsing of
NFS file-systems. These operations are also used to browse server
exports. A v4 server supports export browsing by including exported
directories in a pseudo-file-system. A browsing client can cross
seamlessly between a pseudo-file-system and a real, exported file-
system. Clients must support volatile filehandles and recognize
mount point crossing of server file-systems.

Expires: March 2000 [Page 37]

Draft Protocol Specification NFS version 4 September 1999

8. File Locking

Integrating locking into NFS necessarily causes it to be state-full,
with the invasive nature of "share" file locks it becomes
substantially more dependent on state than the traditional
combination of transport protocols, operating systems NFS and filesystems, simplicity, NLM [XNFS]. There are three components to
making this state manageable:

o Clear division between client and good performance. The NFS version 4
revision has the following goals: server

o Improved access Ability to reliably detect inconsistency in state between client
and good performance on server

o Simple and robust recovery mechanisms

In this model, the Internet. server owns the state information. The protocol client
communicates its view of this state to the server as needed. The
client is designed also able to transit firewalls easily, perform
well where latency detect inconsistent state before modifying a
file.

To support Windows "share" locks, it is high and bandwidth necessary to atomically open
or create files. Having a separate share/unshare operation will not
allow correct implementation of the Windows OpenFile API. In order
to correctly implement share semantics, the existing mechanisms used
when a file is low, and scale opened or created (LOOKUP, CREATE, ACCESS) need to
very large numbers be
replaced. NFS V4 will have an OPEN procedure that subsumes the
functionality of clients per server.

o Strong security with negotiation built into LOOKUP, CREATE, and ACCESS. However, because many
operations require a file handle, the protocol.

The protocol builds traditional LOOKUP is preserved
to map a file name to file handle without establishing state on the work
server. Policy of granting access or modifying files is managed by
the ONCRPC working group in
supporting server based on the RPCSEC_GSS protocol. Additionally NFS version 4
provides a mechanism to allow clients and servers to negotiate
security and require clients and servers client's state. It is believed that these
mechanisms can implement policy ranging from advisory only locking to support
full mandatory locking. While ACCESS is just a minimal
set subset of security schemes.

o Good cross-platform interoperability.

The protocol features OPEN, the
ACCESS procedure is maintained as a filesystem model lighter weight mechanism.

8.1. Definitions

Lock The term "lock" will be used to refer to both record
(byte-range) locks as well as file (share) locks unless
specifically stated otherwise.

Client Throughout this proposal the term "client" is used to
indicate the entity that provides maintains a useful,
common set of features that does not unduly favor locks on behalf
of one filesystem or operating system over another.

o Designed for protocol extensions. more applications. The protocol client is designed to accept standard extensions that do
not compromise backward compatibility. responsible for
crash recovery of those locks it manages. Multiple clients
may share the same transport and multiple clients may exist

Expires: December 1999 March 2000 [Page 6] 38]

Draft Protocol Specification NFS version 4 June September 1999

2. RPC and Security Flavor

2.1. Ports and Transports

Historically, NFS version 2 and version 3 servers have resided

on
UDP/TCP port 2049. Port 2049 is a IANA registered port number for NFS
and therefore will continue the same network node.

Clientid A 64-bit quantity returned by a server that uniquely
corresponds to be used for NFS version 4. Using a client supplied Verifier and ID.

Lease An interval of time defined by the
well known port server for NFS services means which the NFS
client will not need
to use is irrevokeably granted a lock. At the RPC binding protocols as described in [RFC1833]; this will
allow NFS to transit firewalls. end of a
lease period the lock may be revoked if the lease has not
been extended. The NFS lock must be revoked if a conflicting
lock has been granted after the lease interval. All leases
granted by a server SHOULD offer its RPC service via TCP as have the same fixed interval.

Stateid A 64-bit quantity returned by a server that uniquely
defines the locking state granted by the primary
transport. The server SHOULD also provide UDP for RPC service. The
NFS client SHOULD also have a preference
specific lock owner for TCP usage but may supply a mechanism to override TCP in favor specific file. A stateid
composed of UDP as the RPC transport.

2.2. Security Flavors

Traditional RPC implementations all bits 0 or all bits 1 have included AUTH_NONE, AUTH_SYS,
AUTH_DH, special meaning
and AUTH_KRB4 as security flavors. With [RFC2203] an
additional security flavor of RPCSEC_GSS has been introduced which
uses the functionality of GSS-API [RFC2078]. This allows for the use
of varying security mechanisms are reserved.

Verifier A 32-bit quantity generated by the RPC layer without the
additional implementation overhead of adding RPC security flavors.
For NFS version 4, client that the RPCSEC_GSS security flavor MUST be used server
can use to
enable determine if the mandatory security mechanism. The flavors AUTH_NONE,
AUTH_SYS, client has restarted and AUTH_DH MAY be implemented as well.

2.2.1. Security mechanisms for NFS version 4

The use of RPCSEC_GSS requires selection of: mechanism, quality of
protection, lost
all previous lock state.

8.2. Locking

It is assumed that manipulating a lock is rare when compared to I/O
operations. It is also assumed that crashes and service (authentication, integrity, privacy). The
remainder of this document will refer network partitions
are relatively rare. Therefore it is important that I/O operations
have a light weight mechanism to these three parameters of
the RPCSEC_GSS security as indicate if they possess a held
lock. A lock request contains the security triple.

2.2.1.1. Kerberos V5 as security triple

The Kerberos V5 GSS-API mechanism as described in [RFC1964] MUST be
implemented heavy weight information required
to establish a lock and provide uniquely define the lock owner.

The following security triples.

columns:

Expires: December 1999 [Page 7]

Draft Protocol Specification NFS version 4 June 1999

1 == number of pseudo flavor
2 == name of pseudo flavor
3 == mechanism's OID
4 == mechanism's algorithm(s)
5 == RPCSEC_GSS service

This section will be expanded to include sections describe the pertinent
details transition from draft-ietf-nfsv4-nfssec-00.txt.

2.2.1.2. <another security triple>

Another GSS-API mechanism will need to be specified here
along with the corresponding security triple(s).

2.3. Security Negotiation

With heavy weight
information to the NFS version 4 eventual stateid used for most client and server potentially offering multiple security
mechanisms,
locking and lease interactions.

8.2.1. Client ID

For each LOCK request, the client will need must identify itself to the server.
This is done in such a way as to determine or negotiate
which mechanism allow for correct lock
identification and crash recovery. Client identification is
accomplished with two values.

o A verifier that is used to detect client reboots.

o A variable length opaque array to uniquely define a client.

For an operating system this may be used for its communication with the server.
The a fully qualified host

Expires: March 2000 [Page 39]

Draft Protocol Specification NFS server may have multiple points within its file system version 4 September 1999

name
space that are available or IP address, and for use by NFS clients. In turn the a user level NFS
server may be configured such that each of these entry points client it may
have different
additionally contain a process id or multiple security mechanisms in use. other unique sequence.

The security negotiation between client and server must be done with data structure for the Client ID would then appear as:
struct nfs_client_id {
opaque verifier[4];
opaque id<>;
}

It is possible through the mis-configuration of a secure channel to eliminate client or the possibility
existence of a third party
intercepting rogue client that two clients end up using the negotiation sequence and forcing same
nfs_client_id. This situation is avoided by 'negotiating' the
nfs_client_id between client and server to choose a lower level of security than required/desired.

2.3.1. Security Error

Based on with the assumption that each NFS version 4 client and server
must support a minimum set use of security (i.e. Kerberos-V5 under
RPCSEC_GSS, <ed: add other>), the NFS client will start its
communication with
SETCLIENTID. The following describes the server with one two scenarios of
negotiation.

1 Client has never connected to the minimal security

Expires: December 1999 [Page 8]

Draft Protocol Specification NFS version 4 June 1999

triples. During communication with the server, server

In this case the client may
receive generates an NFS error of NFS4ERR_WRONGSEC. This error allows nfs_client_id and
unless another client has the same nfs_client_id.id field,
the server accepts the request. The server also records the
principal (or principal to notify uid mapping) from the client credential
in the RPC request that contains the security triple currently being
used is not appropriate for access nfs_client_id
negotiation request.

Two clients might still use the same nfs_client_id.id due
to perhaps configuration error (say a High Availability
configuration where the server's file system
resources. The client nfs_client_id.id is then responsible for determining what
security triples are available at derived from
the server ethernet controller address and choose one which is
appropriate for both systems have the client.

2.3.2. SECINFO

The new procedure SECINFO (see SECINFO procedure definition) will
allow
same address). In this case, nfs4err can be a switched
union that returns in addition to NFS4ERR_CLID_INUSE, the
network address (the rpcbind netid and universal address)
of the client to determine, on a per filehandle basis, what
security triple that is using the id.

2 Client is re-connecting to be used for the server access. after a client reboot

In general, this case, the client still generates an nfs_client_id
but the nfs_client_id.id field will not have be the same as the
nfs_client_id.id generated prior to reboot. If the server
finds that the principal/uid is equal to use the SECINFO procedure except during
initial communication previously
"registered" nfs_client_id.id, then locks associated with
the server or when old nfs_client_id are immediately released. If the
principal/uid is not equal, then this is a rogue client crosses
policy boundaries at the server. It could happen that the server's
policies change during and
the client's interaction therefore forcing request is returned in error. For more discussion of
crash recovery semantics, see the
client to negotiate a new security triple. section on "Crash
Recovery"

Expires: December 1999 March 2000 [Page 9] 40]

Draft Protocol Specification NFS version 4 June September 1999

3. File handles

The file handle in the NFS protocol

In both cases, upon success, NFS4_OK is a per returned. To help reduce the
amount of data transferred on OPEN and LOCK, the server will also
return a unique identifier
for 64-bit clientid value that is a file system object. The contents of the file handle are opaque short hand reference
to the nfs_client_id values presented by the client. Therefore, From this point
forward, the server is responsible for translating client can use the file handle clientid to an internal representation of the file system
object. Since the file handle is the client's reference refer to an object itself.

8.2.2. nfs_lockowner and the client may cache this reference, the server should not reuse stateid Definition

When requesting a file handle for another file system object. If lock, the server needs client must present to
reuse a file handle value, the time elapsed before reuse SHOULD be
large enough that it is likely the client no longer has a cached copy
of server the reused file handle value.

3.1. Obtaining
clientid and an identifier for the First File Handle

The procedures owner of the NFS protocol requested lock.
These two fields are defined in terms of one or
more file handles. Therefore, the client needs a file handle referred to
initiate communication with as the server. With NFS version 2 [RFC1094] nfs_lockowner and NFS version 3 [RFC1813], there exists an ancillary protocol to
obtain this first file handle. The MOUNT protocol, RPC program
number 100005, provides the mechanism
definition of translating a string based
file system path name to a file handle which can then be used those fields are:

o A clientid returned by the
NFS protocols.

The MOUNT protocol has deficiencies in the area of security and use
via firewalls. This is one reason that the use server as part of the public file
handle was introduced [RFC2054] [RFC2055]. With the clients use of public
file handle in combination with
the LOOKUP SETCLIENTID procedure in NFS version 2
and 3, it has been demonstrated that the MOUNT protocol is
unnecessary for viable interaction between NFS client and server.

Therefore, NFS version 4 will not use an ancillary protocol for
translation from string based path names

o A variable length opaque array used to uniquely define the owner
of a file handle. Two
special file handles will be used as starting points for lock managed by the NFS client.

3.1.1. Root File Handle

The first of

This may be a thread id, process id, or other unique value.

When the special file handles is server grants the ROOT file handle. The
ROOT file handle does not have lock it responds with a special file handle value unique 64-bit
stateid. The stateid is used as does a short hand reference to the
nfs_lockowner, since the public file handle. The ROOT file handle is server will be maintaining the "conceptual"
root
correspondence between them.

8.2.3. Use of the file system name space at stateid

All I/O requests contain a stateid. If the NFS server. The client
uses or starts with nfs_lockowner performs
I/O on a range of bytes within a locked range, the ROOT file handle stateid returned
by employing the PUTROOTFH
procedure. The PUTROOTFH procedure instructs the server must be used to set indicate the
"current" file handle to appropriate lock (record
or share) is held. If no state is established by the ROOT client, either
record lock or share lock, a stateid of the server's file tree. Once
this PUTROOTFH procedure all bits 0 is used, used. If no
conflicting locks are held on the client can then traverse file, the
entirety of server may grant the server's file tree I/O
request. If a conflict with an explicit lock occurs, the LOOKUP procedure. request is
failed (NFS4ERR_LOCKED). This allows "mandatory locking" to be
implemented.

A stateid of all bits 1 allows read requests to bypass locking checks
at the server. However, write requests with stateid with bits all 1
does not bypass file locking requirements.

An explicit lock may not be granted while an I/O operation with
conflicting implicit locking is being performed.

Expires: December 1999 March 2000 [Page 10] 41]

Draft Protocol Specification NFS version 4 June September 1999

complete discussion

The byte range of the server name space a lock is in section 6, "NFS
Server Name Space".

3.1.2. Public File Handle

Unlike the root file handle, the public file handle indivisible. A range may be locked,
unlocked, or changed between read and write but may not have
subranges unlocked or changed between read and write. This is represented the
semantics provided by Win32 but only a reserved or special value subset of the file handle. NFS version 4
defines the file handle as a variable length array semantics
provided by Unix. It is expected that Unix clients can more easily
simulate modifying subranges than Win32 servers adding this feature.

8.2.4. Sequencing of bytes (see
section 4, "Basic Data Types"). The public file handle Lock Requests

Locking is different than most NFS operations as it requires "at-
most-one" semantics that are not provided by ONC RPC. In the 'zero'
file handle face of
retransmission or in other words reordering, lock or unlock requests must have a file handle with
well defined and consistent behavior. To accomplish this each lock
request contains a array length of
zero.

Again unlike the root file handle, the public file handle may be
bound or represent an arbitrary file system object at the server. sequence number that is a monotonically increasing
integer. Different nfs_lockowners have different sequences. The
server is responsible for this binding. It may be that maintains the
public file handle last sequence number (L) received and the root file handle refer to the same file
system object. However,
response that was returned. If a request with a previous sequence
number (r < L) is received it is up to the administrative software at
the server and the policies of the server administrator to define the
binding of public file handle and server file system object. The
client may not make any assumptions about this binding.

3.2. File Handle Types

In NFS version 2 and 3, there silently ignored as its response
must have been received before the last request (L) was one type of file handle with a
single set of semantics. NFS version 4 introduces sent. If a new type of file
handle in an attempt to accommodate certain server environments. The
first type
duplicate of file handle last request (r == L) is 'persistent'. The semantics of received, the stored response
is returned. If a
persistent file handle request beyond the next sequence (r == L + 2) is
received it is silently ignored. Sequences are reinitialized
whenever the same as client verifier changes.

8.3. Blocking Locks

Some clients require the file handles support of NFS
version 2 and 3. blocking locks. The second or new type of file handle is current
proposal lacks a call-back mechanism, similar to NLM, to notify a
client when the
'volatile' file handle.

The volatile file handle type is being introduced lock has been granted. Clients have no choice but to address server
functionality or implementation issues
continually poll for the lock, which prevent correct or
feasible implementation of a persistent file handle. Some server
environments do not provide presents a file system level invariant that can be fairness problem.
Two new lock types are added, READW and WRITEW used to construct indicate to
the server that the client is requesting a persistent file handle. blocking lock. The underlying server
file system may not provide
should maintain an ordered list of pending blocking locks. When the invariant or
conflicting lock is released, the server's file system
APIs server may not provide access to wait the needed invariant. Volatile file
handles may ease lease period
for the implementation of server functionality such as
hierarchical storage management or file system reorganization or
migration. However, first client to re-request the volatile file handle increases lock. After the
implementation burden for lease period
expires the next waiting client but this increased burden request is
deemed acceptable based on the overall gains achieved by allowed the
protocol.

Since lock. Clients
are required to poll at an interval sufficiently small that it is
likely to acquire the client will have different paths of logic lock in a timely manner. The server is not
required to handle
persistent and volatile file handles, maintain a file attribute list of pending blocked locks as it is defined used to
increase fairness and not correct operation. Because of the
unordered nature of crash recovery, storing of lock state to stable
storage would be required to guarantee ordered granting of blocking
locks.

Expires: December 1999 March 2000 [Page 11] 42]

Draft Protocol Specification NFS version 4 June September 1999

which

8.4. Lease Renewal

The purpose of a lease is to allow a server to remove stale locks
that are held by a client that has crashed or is otherwise
unreachable. It is not a mechanism for cache consistency and lease
renewals may not be used by denied if the client to determine lease interval has not expired.
Any I/O request that has been made with a valid stateid is a positive
indication that the file handle types client is still alive and locks are being returned by the server.

3.2.1. General Properties
maintained. This becomes an implicit renewal of a File Handle

The file handle contains all the information lease. In the server needs to
distinguish an individual file. To
case no I/O has been performed within the client, lease interval, a lease can
be renewed by having the file handle is
opaque. The client stores file handles for use in issue a later request and
can compare two file handles from zero length READ. Because
the same server nfs_lockowner contains a unique client value, any stateid for equality by
doing a byte-by-byte comparison, but MUST NOT otherwise interpret the
contents of file handles. If two file handles from
client will renew all leases for locks held with the same server
are equal, they MUST refer to client
field. This will allow very low overhead lease renewal that scales
extremely well. In the same file, but if they are not
equal, typical case, no conclusions can be drawn. Servers SHOULD try to maintain a
one-to-one correspondence between file handles extra RPC calls are needed
and files but this is
not required. Clients MUST only use file handle comparisons only to
improve performance, not for correct behavior.

As an example, in the worst case that two different path names when
traversed at the server terminate at the same file system object, one RPC is required every lease period
regardless of the
server SHOULD return number of locks held by the same file handle for each path. This can
occur if a hard link client.

8.5. Crash Recovery

The important requirement in crash recovery is used to create two file names which refer to that both the same underlying file object and associated data. For example, if
paths /a/b/c client
and /a/d/c refer to the same file, the server SHOULD
return know when the same file handle for both path names traversals.

3.2.2. Persistent File Handle

A persistent file handle other has failed. Additionally it is defined as having
required that a persistent value for
the lifetime client sees a consistent view of data across server
reboots. All I/O operations that may have been queued within the file system object to which it refers. Once
client or network buffers must wait until the
server creates client has successfully
recovered the file handle for a file system object, locks protecting the server
MUST return I/O operations.

8.5.1. Client Failure and Recovery

In the same file handle for event that a client fails, the object for server may recover the lifetime of client's
locks when the object. associated leases have expired. Conflicting locks
from another client may only be granted after this lease expiration.
If the server restarts client is able to restart or reboots, reinitialize within the NFS server must
honor and present lease
period the same file handle value as it did client may be forced to wait the remainder of the lease
period before obtaining new locks.

To minimize client delay upon restart, lock requests contain a
verifier field in the
server's previous instantiation.

The persistent file handle lock_owner. This verifier is part of the
initial SETCLIENTID call made by the client. Since the verifier will
be become stale or invalid when changed by the
file system object is removed. When client upon each initialization, the server is presented with can
compare a
persistent file handle that refers new verifier to a deleted object, it MUST
return an error of NFS4ERR_STALE. A file handle may become stale
when the file system containing the object is no longer available.
The file system may become unavailable if it exists on removable
media verifier associated with currently
held locks and determine that they do not match. This signifies the media is no longer available at
client's new instantiation and loss of locking state. As a result,
the server or the file
system in whole has been destroyed or the file system has simply been
removed from is free to release all locks held which are associated
with the server's name space (i.e. unmounted in a Unix
environment). old verifier.

Expires: December 1999 March 2000 [Page 12] 43]

Draft Protocol Specification NFS version 4 June September 1999

3.2.3. Volatile File Handle

A volatile file handle does not share

For secure environments, a change in the same longevity attributes verifier must only cause the
release of locks associated with the persistent file handle. The server may determine that a
volatile file handle authenticated requester. This
is no longer required to prevent a rogue entity from freeing otherwise valid at many different points in
time.
locks. Note that the verifier must have the same uniqueness
properties of the COMMIT verifier.

8.5.2. Server Failure and Recovery

If the server can definitively determine that a volatile file
handle refers to an object that has been removed, fails and loses locking state, the server should
return NFS4ERR_STALE must wait the
lease period before granting any new locks or allowing any I/O. An
I/O request during the grace period with a stale stateid will fail
with NFS4ERR_GRACE. To recover the lock and associate state, the
client will reissue the lock request with reclaim set to TRUE. Upon
receiving a successful reply and associated stateid, the client (as is may
reissue the case for persistent
file handles). In all other cases where I/O request with the server determines that new stateid.

Any time a
volatile file handle can no longer be used, it should return client receives an NFS4ERR_GRACE error, the client must
assume that all locking state associated with the server returning
the error
of NFS4ERR_EXPIRED. has been lost. The following table shows client should start recovering all
outstanding locks upon receiving NFS4ERR_GRACE.

If the most common points at which server receives a volatile
file handle may expire. This table represents the view from lock request during its grace period that
does not have reclaim set to TRUE, the
client's perspective and as such provides columns for when server must return
NFS4ERR_GRACE. This error return will trigger the file
may be open or closed client to recover
all of its locking state by reclaiming locks.

A lock request outside the client.

Server Provides Persistent or Volatile File Handle
File Open File Closed
___________________________________________________________________
Restart of Server (note 4) P / V P / V
Fileset Migration (note 5) P / V P / V
SHARE/LOCK recovery P / V N/A (note 1)
Client RENAMEs object P / V P / V
Client RENAMEs path to object P / V P / V
Other client RENAMEs object P / V P / V
Other client RENAMEs path server's grace period with reclaim set to object P / V P / V
Client REMOVEs object P / V (note 2) P / V
Other client REMOVEs object P / V N/A (note 3)

Note 1
TRUE can only succeed if the server can guarantee that no conflicting
lock or I/O request has been granted since reboot.

8.5.3. Network Partitions and Recovery

If the file is not open, persistence duration of a network partition is greater than the lease
period provided by the server, the file handle is server will have not
applicable for received a
lease renewal from the recovery of SHARE/LOCK.

Note 2
With NFS version 2 and 3, when client. If this occurs, the client removes server may free
all locks held for the client. As a file it has
open it follows result, all stateids held by the convention of RENAMEing
client will become invalid. Once the file client is able to '.nfsXXXX'
until reach the file is closed. At this point
server after such a network partition, all I/O submitted by the REMOVE is done at
client with the
server.

If now invalid stateids will fail with the server
returning the error NFS4ERR_EXPIRED. Once this same model error is used for v4 then this entry received,
the client will be
'N/A'.

Note 3
If suitably notify the file is not open by application that held the client, then it should not expect
any cached file handle lock.

As a courtesy to be valid. the client or optimization, the server may continue
to hold locks on behalf of a client for which recent communication
has extended beyond the lease period. If the server receives a lock
or I/O request that conflicts with one of these courtesy locks, the

Expires: December 1999 March 2000 [Page 13] 44]

Draft Protocol Specification NFS version 4 June September 1999

Note 4
The restart

server must free the courtesy lock and grant the new request.

In the event of a network partition with a duration extending beyond
the expiration of a client's leases, the NFS server signifies when MUST employ a method
of recording this fact in its stable storage. Conflicting locks
requests from another client may be serviced after the operating system lease
expiration. There are various scenarios involving server failure
after such an event that require the storage of these lease
expirations or NFS software network partitions. One scenario is (re)started. This also includes High
Availability configurations where as follows:

A client holds a separate operating system
instantiation acquires ownership of lock at the file system resources server and encounters a
network resources (i.e. disks partition and IP addresses).

Note 5
Fileset migration occurs when is unable to renew the associated
lease. A second client obtains a conflicting lock and then
frees the lock. After the unlock request by the second
client, the ownership of file system
resources are transfered from one server to another without a
transfer of ownership of reboots or reinitializes. Once the
server recovers, the network resources. So partition heals and the
original client attempts to reclaim the original lock.

In this differs
from scenario and without any state information, the High Availability scenario. This also includes server will
allow the move
of a file system resources within reclaim and the client will be in an inconsistent state
because the same server such that or the
fsid value is different.

The fileset migration entry is a place holder until a file
set migration proposal client has been fully evaluated and decided
upon.

3.2.4. One Method no knowledge of Constructing a Volatile File Handle

As mentioned, the conflicting
lock.

The server may choose to store this lease expiration or network
partitioning state in some instances a file handle is stale (no longer
valid, perhaps because way that will only identify the file was removed from client as a
whole. Note that this may potentially lead to lock reclaims being
denied unnecessarily because of a mix of conflicting and non-
conflicting locks. The server may also choose to store information
about each lock that has an expired lease with an associated
conflicting lock. The choice of the server), or it amount and type of state
information that is expired (the underlying file stored is valid, but since left to the file handle
is volatile, it may have expired). Thus implementor. In any case,
the server needs to be able must have enough state information to return NFS4ERR_STALE in the former case, enable correct
recovery from multiple partitions and NFS4ERR_EXPIRED in multiple server failures.

8.6. Server Revocation of Locks

At any point, the latter case. This server can be done revoke locks held by careful construction of a client and the
volatile file handle. One possible implementation follows.

A volatile file handle, while opaque to
client must be prepared for this event. When the client could contain:

[volatile bit = 1 | server boot time | slot | generation number]

o slot is an index in detects that
its locks have been or may have been revoked, the server volatile file handle table

o generation number client is the generation number
responsible for validating the table
entry/slot

If state information between itself and
the server boot time is less than server. Validating locking state for the current server boot time,
return NFS4ERR_EXPIRED. If slot is out client means that it
must verify or reclaim state for each lock currently held.

The first instance of range, return
NFS4ERR_EXPIRED. If lock revocation is upon server reboot or re-
initialization. In this instance the generation number does not match, return
NFS4ERR_EXPIRED.

When client will receive an error or
NFS4ERR_GRACE and the server reboots, client will proceed with normal crash recovery
as described in the table is gone (it is volatile). previous section.

Expires: December 1999 March 2000 [Page 14] 45]

Draft Protocol Specification NFS version 4 June September 1999

If volatile bit

The second lock revocation event can occur as a result of
administrative intervention within the lease period. While this is 0, then
considered a rare event, it is possible that the server's
administrator has decided to release or revoke a persistent file handle with particular lock held
by the client. As a
different structure following it.

3.3. Client Recovery from File Handle Expiration

With result of revocation, the introduction client will receive an
error of NFS4ERR_EXPIRED and the volatile file handle, error is received within the lease
period for the lock. In this instance the client must
take on additional responsibility so may assume that it
only the lock_owner's locks have been lost. The client notifies the
lock holder appropriately. The client may prepare itself to
recover from not assume the expiration of lease
period has been renewed as a volatile file handle. If result of failed operation.

The third lock revocation event is the server inability to renew the lease
period. While this is return persistent file handles, considered a rare or unusual event, the client does not need these
additional steps.

For volatile file handles, most commonly
must be prepared to recover. Both the server and client will need be able
to
store detect the component names leading up failure to renew the lease and including are capable of
recovering without data corruption. For the file system
object in question. With these names, server, it tracks the client should be able to
recover by finding a file handle in
last renewal event serviced for the name space that is still
available or by starting at client and knows when the root of lease
will expire. Similarly, the server's file system name
space.

If client must track operations which will
renew the expired file handle refers lease period and is able to an object determine lease period
expiration.

When the client determines that the lease period has been removed
from expired, the file system, obviously
client must mark all locks held for the associated lease as
"unvalidated". This means the client will not be able has been unable to
recover from re-establish
or confirm the expired file handle.

It is also possible that appropriate lock state with the server. As described
in the previous section on crash recovery, there are scenarios in
which the server may grant conflicting locks after the lease period
has expired file handle, refers to for a file client. Once the lease period has expired, the
client must validate each lock it has held to ensure that a
conflicting lock has not been renamed. granted. The client may accomplish
this task by issuing an I/O request, either a pending I/O or zero
length read. If the file was renamed by another client,
again it response to the request is possible that success, the original client will not be able to
recover. However, in
has validated the case that lock and re-established the client appropriate state
between itself and the server. If the I/O request is renaming not successful,
the
file lock was revoked by the server and the file client must notify the
owner.

8.7. Share Reservations

A share reservation is open, it a mechanism to control access to a file. It
is possible that the a separate and independent mechanism from record locking. When a
client may be able opens a file, it issues an OPEN request to recover. The client can determine the new path name based on server
specifying the
processing type of access required (READ, WRITE, or BOTH) and the rename request. The client can then regenerate
type of access to deny others (deny NONE, READ, WRITE, or BOTH). If
the
new file handle based on OPEN fails the new path name. The client could also
use will fail the compound operation mechanism to construct a set applications open request.

Pseudo-code definition of operations
like:
RENAME A B
LOOKUP B
GETFH the semantics:

Expires: December 1999 March 2000 [Page 15] 46]

Draft Protocol Specification NFS version 4 June September 1999

4. Basic Data Types

Arguments and results from operations will be described in terms of
basic XDR types defined in [RFC1832]. The following data types will
be defined in terms of basic XDR types:

filehandle: opaque <128>

An NFS version 4 filehandle. A filehandle with zero length is
recognized as a "public" filehandle.

utf8string: opaque <>

A counted array of octets that contains

if ((request.access & file_state.deny)) ||
(request.deny & file_state.access))
return (NFS4ERR_DENIED)

8.8. OPEN/CLOSE Procedures

To provide correct share semantics, a UTF-8 string.

Note: Section 11, Internationalization, covers client MUST use the rational of
using UTF-8.

bitmap: uint32 <>

A counted array of 32 bit integers used OPEN
procedure to contain bit values.
The position obtain the initial file handle and indicate the desired
access and what if any access to deny. Even if the client intends to
use a stateid of all 0's or all 1's, it must still obtain the integer in
filehandle for the array that contains bit n can
be computed from regular file with the expression (n / 32) and its bit within OPEN procedure. For clients
that
integer is (n mod 32).

0 1
+-----------+-----------+-----------+--
| count | 31 .. 0 | 63 .. 32 |
+-----------+-----------+-----------+--

createverf: opaque<8>

Verify used do not have a deny mode built into their open API, deny equal to
NONE should be used.

The OPEN procedure with the CREATE flag, also subsumes the CREATE
procedure for exclusive regular files as used in previous versions of NFS,
allowing a create with a share to be done atomicly.

Will expand on create semantics

nfstime4
struct nfstime4 {
int64_t seconds;
uint32_t nseconds;
} here.

The nfstime4 structure gives CLOSE procedure removes all share locks held by the number lock_owner on
that file. If record locks are held they should be explicitly
unlocked. Some servers may not support the CLOSE of seconds a file that
still has record locks held; if so, CLOSE will fail and
nanoseconds since midnight or 0 hour January 1, 1970 Coordinated
Universal Time (UTC). Values greater than zero for the seconds
field denote dates after the 0 hour January 1, 1970. Values
less than zero for the seconds field denote dates before return an
error.

The LOOKUP procedure is preserved and will return a file handle
without establishing any lock state on the 0
hour January 1, 1970. In both cases, server. Without a valid
stateid, the nseconds field is to
be added to server will assume the seconds field for client has the final time representation. least access. For
example, if the time to a file opened with deny READ/WRITE cannot be represented is one-half second accessed using
a file handle obtained through LOOKUP.

Expires: December 1999 March 2000 [Page 16] 47]

Draft Protocol Specification NFS version 4 June September 1999

before 0 hour January 1, 1970, the seconds field would have a
value

9. Client-Side Caching

Client-side caching of negative one (-1) data, of file attributes, and the nseconds fields would have of file names is
essential to providing good performance in NFS. Providing dis-
tributed cache-coherence is a
value difficult problem and previous versions
of one-half second (500000000). Values greater than
999,999,999 NFS have not attempted it. Instead, several client implementation
techniques have been used to reduce the problems that lack of co-
herence poses for nseconds are considered invalid.

This data type users. These techniques have not been clearly
defined by earlier specifications and it is often unclear what is
valid or invalid client behavior.

NFS version 4 uses many techniques similar to those that have been
used in previous versions of NFS. It does not provide distributed
cache coherence, but it defines a more limited set of caching
guarantees to pass time allow locks and date information. A
server converts share reservation to and be used without
destructive interference from local time when processing time
values, preserving as much accuracy as possible. If client-side caching.

In addition, version 4 introduces a delegation mechanism which allows
many decisions normally made by the
precision server to be made locally by
clients. This provides efficient support of timestamps stored the common cases where
sharing is infrequent or where sharing is read-only.

9.1. Performance Challenges for Client-Side Caching

Caching techniques used in previous versions of NFS have been
successful in providing good performance. However, several scala-
bility challenges can arise when those techniques are used with very
large numbers of clients, particularly when those clients are
geographically distributed, increasing the latency for cache
revalidation requests.

When latencies are large, repeated cache validation requests at open
time, which NFS-v2 and NFS-v3 clients typically do, can have serious
performance drawbacks. A common case is one in which a file system object is less
than defined, loss of precision can occur. An adjunct time
maintenance protocol only
accessed by a single client. Sharing is infrequent.

In this case, repeated reference to the server to find that no
conflicts exist, is expensive. A better option is recommended to reduce allow a client
repeatedly opening a file to do so without reference to the server,
until potentially conflicting operations from another client actually
occur.

A similar situation arises in connection with file locking. Sending
file lock and unlock requests to the server
time skew.

specdata4
struct specdata4 {
uint32_t specdata1;
uint32_t specdata2;
}

This as well as the I/O
requests necessary to make data type represents additional information for caching consistent with the device
file types NFCHR locking
semantics (see the section "Data Caching and NFBLK. File Locking") can
severely limit performance. When locking is used to provide pro-

Expires: December 1999 March 2000 [Page 17] 48]

Draft Protocol Specification NFS version 4 June September 1999

5. File Attributes

To meet

tection against infrequent conflicts, a large penalty will be paid,
which may discourage the use of locking.

In NFS Version 4 requirements of extensibility and increased
interoperability with non-Unix platforms, attributes must 4, more aggressive caching strategies are designed:

o To be handled
in compatible with a more flexible manner. The large range of server semantics.

o Provide the same caching benefits as previous versions of NFS Version 3 fattr3 structure
contained
when unable to provide the more aggressive model.

o Requirements for aggressive caching are organized so that a fixed list
large portion of attributes that the benefit can be obtained even when not all clients and servers
are able to support or care about, which cannot be extended as new
needs crop up, and which provides no way to indicate non-support.
With NFS Version 4,
of the client will requirements can be able to ask what attributes met.

The appropriate requirements for the server supports, and will be able to request only those
attributes in which it is interested.

To this end, attributes will be divided into three groups: mandatory,
recommended and named. Both mandatory and recommended attributes are
supported discussed in the NFS V4 protocol by a later
sections in which specific and well-defined
encoding, and are identified by number. They forms of caching are requested by
setting a bit in the bit vector sent in dealt with. (see
Section "Open Delegation").

NOTE: [[This discussion of proxy caching assumes that the GETATTR request; a
proxy server appears to the (real) server response includes as an ordinary
client. Should there be a bit vector to list what attributes were
returned in response. New mandatory or recommended attributes may proposal for non-transparent
proxy server support (Mike Eisler's proxy model 2), this
can be
added to the NFS protocol between revisions by publishing modified.]]

9.2. Proxy Caching

Proxy caching is a
standards-track RFC which allocates useful technique to reduce latency and avoid
server overload when a new attribute large number value and
defines the encoding for the attribute.

Named attributes are accessed of geographically distributed
clients share data. The proxy cache allows many requests to be
satisfied by the new OPENATTR operation, which
accesses a hidden directory of attributes local server, reducing bandwidth and latencies
associated with a
filesystem object. OPENATTR takes a filehandle for the object and
returns the filehandle for the attribute hierarchy, which is a
directory object accessible by LOOKUP or READDIR, and which contains
files whose names represent the named attributes and whose data bytes
are the value of accessing the attribute. For example:

LOOKUP "foo" ; look up file
GETATTR attrbits
OPENATTR ; access foo's named attributes
LOOKUP "x11icon" ; look up specific attribute
READ 0,4096 ; read stream of bytes

Named attributes are intended primarily for data needed by
applications rather than by an NFS client implementation per se; primary server.

If NFS
implementors are strongly encouraged version 4 were to define their new attributes
as recommended attributes by bringing them limit itself to the working group.

The set caching approaches used
in NFS v2 and NFS v3, a large number of attributes the requests which are classified as mandatory is
deliberately small, since servers must do whatever it takes to
support them. The recommended attributes may be unsupported, though a proxy
server should support as many as it can. Attributes are deemed would receive would result in corresponding requests to the
distant server:

o All OPEN and CLOSE requests

o WRITE requests necessary to flush out dirty data before all file
close operations.

o All LOCK and UNLOCK requests.

Expires: December 1999 March 2000 [Page 18] 49]

Draft Protocol Specification NFS version 4 June September 1999

mandatory if

o READ and WRITE requests which must go to the data server because
locks are held or being released.

o All directory modification requests (e.g. CREATE, REMOVE, etc.)

o All SETATTR requests

o Many other requests because of cache entry staleness

Maintaining distributed caches allowing authoritative decisions to be
made locally is both needed by difficult, in the general case. However, there are
many situations in which access patterns allow such decisions to be
delegated opportunistically to particular clients (such as proxy
servers) avoiding a large number great deal of clients and unnecessary communication. This is not otherwise reasonably computable by the client
of particular importance when support is
not provided on the server.

5.1. Mandatory Attributes

These MUST be supported by every NFS Version 4 client scaling to very large numbers of
clients.

9.3. Delegation and Callbacks

Recallable delegation of server in
order responsibilities for a file to ensure a minimum level
client (which may include proxy servers) improves performance by
avoiding repeated requests to the server in the absence of interoperability. The
interclient conflict. A server must
store and return these attributes, and recalls delegated responsibilities,
using a callback rpc from the server to the client, when another
client must be able engages in sharing of a delegated file.

A delegation is passed from the server to
function with an attribute set limited the client, specifying the
object for which the delegation is being done and type of delegation.
There are different types of delegations but each contains a stateid
to these attributes, though
some be used to represent the delegation when performing operations
that depend on the delegation. This stateid is similar to those
associated with locks and share reservations but differs in that the
stateid for a delegation is associated with a clientid and may be impaired or limited in some ways in this case.
used on behalf of all the nfs_lockowner's for the given client. A
delegation is made to the client as a whole and not to any specific
process within it.

Because callback rpc's may ask not work in all environments (due to
firewalls, for any example), correct operation does not depend on them.
Preliminary testing of these attributes to be returned callback functionality by
setting means of a bit in CB_NULL
request determines whether callbacks can be supported. The CB_NULL
request checks the GETATTR request, and continuity of the callback path. A server must return
their value.

5.2. Recommended Attributes

These attributes are understood well enough makes a
preliminary assessment of callback availability to warrant support in a given client and
avoids delegating responsibilities until it has determined that
callbacks are supported. Because client requests for delegation are
always conditional upon the absence of conflicting access, clients

Expires: March 2000 [Page 50]

Draft Protocol Specification NFS Version version 4 protocol, though they may September 1999

can not assume that a request for delegation will be supported on granted, and
must always be prepared for denial.

Once granted, a delegation behaves in most ways like a lock. There
is an associated lease that is subject to renewal together with all
clients and servers. A client may ask for any
of these attributes to
be returned the other leases held by setting that client.

Unlike locks, a bit in request to a delegated file from a second client will
cause the GETATTR request, but must be able server to deal with not receiving them. A recall a delegation through a callback.

On recall, the client may ask for holding the set of
attributes delegation must flush modified
state (such as modified data) to the server supports and should not request attributes return the
server does
delegation. The conflicting request will not support. A server should be tolerant of requests for
unsupported attributes, and simply not return them, rather than
considering responded to until
the request an error. It recall is expected that servers will
support all attributes they comfortably can, complete, either by the return of the delegation or by
the server timing out the recall and only fail revoking the delegation.
Following recall, the server has the information necessary to support
attributes which are difficult grant
or deny second client's request.

Since recalling a delegation may involve the flushing of substantial
state to support in their operating
environments. A the server, the server should provide attributes whenever they don't
have to "tell lies" allow a time to complete the client. For example,
recall substantially longer than for a file modification typical single RPC. The
server may also extend the time should be either an accurate allowed if it can determine that
state is being diligently flushed by the client. However, the time or
to complete the recall should not be supported by unbounded.

For example, when responsibility to mediate opens on a given file is
delegated to a client (see the server. This section "Open Delegation"), the server
will not always be comfortable to clients but it
seems that know what opens are in effect on the client has a better ability to fabricate or construct
an attribute or do without.

Most attributes from NFS V3's FSINFO, FSSTAT and PATHCONF procedures
have been added as recommended attributes, so that filesystem info
may thus will be collected via
unable to determine whether the filehandle of access and deny state for the file
allows any object particular open until the filesystem.
This renders those procedures unnecessary delegation has been returned.

Client failure or a network partition can result in NFS V4. If failure to
respond to a recall callback. The server
supports any per-filesystem attributes, it must support will revoke the fsid
attribute so that delegation,
rendering any modified state still on the client may always determine when filesystems useless.

9.3.1. Delegation Recovery

There are crossed so three situations that it can work correctly with these attributes. delegation recovery must deal with:

o Client reboot

o Server reboot

o Network partition (full or callback-only)

Expires: December 1999 March 2000 [Page 19] 51]

Draft Protocol Specification NFS version 4 June September 1999

5.3. Named Attributes

These attributes are not supported by direct encoding in

In the NFS
Version 4 protocol but are accessed by string names rather than
numbers and correspond to an uninterpreted stream even of bytes which are
stored with the filesystem object. The namespace for these
attributes may be accessed by using the OPENATTR operation to get a
filehandle for a virtual "attribute directory" and using READDIR and
LOOKUP operations on this filehandle. Named attributes may then be
examined or changed by normal READ and WRITE and CREATE operations on
the filehandles returned from READDIR and LOOKUP. Named attributes
may have attributes, for example, a security label may have access
control information in its own right.

It is recommended that servers support arbitrary named attributes. A client should not depend on reboot, the ability failure to store any named attributes renew leases will
result in the server's filesystem. If a server does support named
attributes, revocation of record locks and share reservations.
Delegations, however, may treated a client which is also able to handle them should bit differently.

Because data associated with some delegations may be able written to copy a file's data
stable storage on the client and meta-data with complete transparency from
one location because a delegation held by a proxy
server may be further delegated to another; this would imply that its client in turn whereupon the
proxy server may reboot, there should will be no
attribute names situations in which
delegations will need to be considered illegal by re-established after a client (which
includes a proxy server) reboots.

To accommodate such situations, the server.

Names of attributes will not be controlled by server may, after leases expire,
force requests that conflict with existing delegations to wait for a standards body.
However, vendors and application writers are encouraged
longer period of time. This is consistent with the fact that recall,
including the time necessary to flush modified state to register
attribute names and the interpretation server
and semantics of return the stream of
bytes via informational RFC so that vendors delegation, may interoperate where
common interests exist.

5.4. Mandatory Attributes - Definitions

Name # DataType Access Description
__________________________________________________________________
supp_attr 0 bitmap READ
The bit vector which take significant time. This longer
interval would retrieve all
mandatory and
recommended attributes allow clients which may be requested
for this object.

The client must ask
this question reboot to consult stable storage
and request correct
attributes.

Expires: December 1999 [Page 20]

Draft Protocol Specification NFS version 4 June 1999

object_type 1 nfs4_ftype READ
The the reclamation of delegations which have not been timed
out using this longer interval. For open delegations, such
delegations are reclaimed using OPEN with a claim type of
CLAIM_DELEGATE_PREV. (See the object
(file, directory,
symlink)

The client cannot
handle object
correctly without
type.

persistent_fh 2 boolean READ
Is the filehandle Sections on "Data Caching and
Revocation" and "Procedure 17: OPEN" for
this object
persistent?

Server should know if discussion of open
delegation and the file handles being
provided are
persistent or not. If details of OPEN respectively).

When the server reboots, delegations are reclaimed (using OPEN with
CLAIM_DELEGATE_PREV) in a similar fashion to record locks and share
reservations. However, there is a slight semantic difference.
Normally, the server decides that a delegation should not able
to make be granted,
it performs the requested action (e.g. OPEN) without granting any
delegation. When this
determination, then happens as part of reclaim, the server grants
the delegation but marks it
can choose volatile or
non-persistent.

change 3 uint64 READ
A value created specially so that the client treats the
delegation as having been granted but recalled by the server so that
it then has the client
can use duty to determine
if a file data,
directory contents or
attributes have been
modified. The write all modified state to the server
can just and
then return the
file mtime in this
field though if delegation. This handling of delegation reclaim
reconciles three principles of NFS Version 4:

o That upon reclaim, a more
precise value exists
then client faithfully reporting resources
assigned to it can be
substituted, for by an earlier server instance, a sequence
number.

Necessary for any
useful caching, likely
to must be available.

object_size 4 uint64 R/W
The size of granted
those resources.

o That the object
in bytes.

Could server has untrammeled authority to determine whether
delegations are to be very
expensive granted and, once granted, whether they
are to derive,
likely be continued.

o That the use of callbacks is not to be depended upon until the
client has proved its ability to be
available. receive them.

Expires: December 1999 March 2000 [Page 21] 52]

Draft Protocol Specification NFS version 4 June September 1999

link_support 5 boolean READ
Does

When a network partition occurs, delegations, like locks and share
reservations will be subject to freeing when the object's
filesystem supports
hard links?

Server can easily
determine if links are
supported.

symlink_support 6 boolean READ
Does lease renewal period
expires, although the object's
filesystem supports
symbolic links?

Server can easily
determine if links server will normally extend the period in which
conflicting requests are
supported.

named_attr 7 boolean READ
Does this object have
named attributes?

fsid.major 8 uint64 READ
Unique filesystem
identifier for held off in the
filesystem holding
this object.

fsid.minor 9 uint64 READ
Unique filesystem
identifier within case of delegations.
Eventually, however, the
fsid.major filesystem
identifier for occurrence of a conflicting request from
another client will cause revocation of the
filesystem holding
this object.

5.5. Recommended Attributes - Definitions

Name # Data Type Access Description
___________________________________________________________________
ACL 10 nfsacl4 R/W
The access control
list for delegation. A blockage
of the object.
[The nature callback (e.g. by later network configuration change) will
have the same effect. A recall request will fail and format revocation of ACLs is still to be
determined.]

Expires: December 1999 [Page 22]

Draft Protocol Specification NFS version 4 June 1999

archive 11 boolean R/W
Whether or not this
file has been archived
since
the time delegation will result.

A client normally finds out about revocation of last
modification
(deprecated a delegation when it
uses a stateid associated with a delegation and receives the error
NFS4ERR_EXPIRED. It also may find out about delegation revocation
after a client reboot when it attempts to reclaim a delegation and
receives that same error. Note that in favor the case of backup_time).

cansettime 12 boolean READ
Whether or not this
object's filesystem
can fill in a revoked write
open delegation, there are issues because data may have been modified
by the times
on client whose delegation is revoked and separately by other
clients. See the section "Revocation Recovery for Write Open
Delegation" for a SETATTR request
without an explicit
time.

case_insensitive 13 boolean READ
Are filename
comparisons on this
filesystem case
insensitive?

case_preserving 14 boolean READ
Is filename discussion of such issues. Note also that when
delegations are revoked information about the revoked delegation will
be written by the server to stable storage (as described in section
7.5) to deal with the case on
this filesystem
preserved?

chown_restricted 15 boolean READ
Will in which a request server reboots after revoking a
delegation but before revoked delegate find out about the revocation.

9.4. Data Caching

When programs share access to a set of files they need to
change ownership be
honored?

filehandle 16 nfs4_fh READ
The filehandle
implemented so as to take account of this
object (primarily for
readdir requests).

fileid 17 uint64 READ
A number uniquely
identifying the file
within possibility of conflicting
access by another program. This is true whether the filesystem.

files_avail 18 uint64 READ
File slots available
to this user programs in
question are on different hosts or reside on the
filesystem containing
this object - this
should same host.

Share reservations and record locks are the facilities that NFS v4
provides to allow programs to co-ordinate access by providing mutual
exclusion facilities. NFS v4 data caching must be implemented so
that it does not vitiate the smallest
relevant limit.

files_free 19 uint64 READ
Free file slots on assumptions that those using these
facilities depend on.

9.4.1. Data Caching and OPENs

In order to avoid invalidating the
filesystem containing
this object - this sharing assumptions that
applications rely on, NFS v4 clients should not provide cached data
to applications or modify it on behalf of an application when it
would not be the smallest
relevant limit. valid to obtain/modify that same data via a READ or
WRITE rpc.

Expires: December 1999 March 2000 [Page 23] 53]

Draft Protocol Specification NFS version 4 June September 1999

files_total 20 uint64 READ
Total file slots

Further, in the absence of open delegation (see the Section "Open
delegation"), two further rules apply. These rules are obeyed in
practice by many NFS v2 and NFS v3 clients.

o The first rule is that cached data present on a client must be
revalidated after doing an OPEN, to make sure that the data for
the filesystem
containing this
object.

hidden 21 boolean R/W
Is file considered
hidden?

homogeneous 22 boolean READ
Whether or not this
object's filesystem is
homogeneous, i.e.
whether pathconf in question, is still validly reflected in the client's
cache. This must be done at least when a client open includes
DENY=WRITE or BOTH, terminating a period in which other clients
may have had the opportunity to open the same for all
filesystem objects.

maxfilesize 23 uint64 READ
Maximum supported file
size with WRITE access.
Clients may choose to do the revalidation more often (i.e. on
opens specifying DENY=NONE) to parallel NFS v3 practice for the
filesystem
benefit of users assuming this
object.

maxlink 24 uint32 READ
Maximum number degree of
links for this object.

maxname 25 uint32 READ
Maximum filename size
supported for this
object.

maxread 26 uint64 READ
Maximum read size
supported for this
object.

maxwrite 27 uint64 READ
Maximum write size
supported cache revalidation.

o The second rule, complementary to the first, is that modified
data must be flushed to the server before closing a file opened
for write. If this
object. rule is not adhered to, the revalidation
done after client OPEN's cannot achieve its purpose. This
attribute SHOULD data
must be
supported if committed to stable storage before the CLOSE is done
since retransmission of the data after a server reboot might not
be possible, once the file is writable. Lack closed.

9.4.2. Data Caching and File Locking

When users do not use share reservations to exclude inconsistent
access, but use file locking instead, there is an analogous set of
this attribute can
lead
constraints that apply to the client
either wasting
bandwidth or not
receiving side data caching. These rules are
effective only if file locking is used in a way which is congruent
with the best
performance.

mime_type 28 utf8<> R/W
MIME body type/subtype actual IO operations being done, as opposed to being used in
a purely conventional way. For example, it is possible to manipulate
a 2MB file, dividing the file into two 1MB regions, and using a lock
for write on byte 0 of the file to represent the right to do IO to
the first region and a lock for write to byte 1 of the file to
represent the right to do IO on the second region. As long as all
applications manipulating the file obey this object.

Expires: December 1999 convention, they will
work on a local file system, but they may not work on NFS v4 unless
clients refrain from data caching.

The first rule is that when a client locks a region, it must
revalidate its data cache if it has any cached data in the region
newly locked and invalidate it if the change attribute shows that the
file may have been written since that data was obtained. (A client
might choose to invalidate all of non-modified cached data that it
has, but invalidating all of the data in the newly locked region is
necessary for correct operation).

Expires: March 2000 [Page 24] 54]

Draft Protocol Specification NFS version 4 June September 1999

mode 29 uint32 R/W
Unix-style permission
bits

The second rule is that before releasing a write lock for this object
(deprecated in favor
of ACLs)

no_trunc 30 boolean READ
If a name longer than
name_max is used, will
an error region,
all modified data for that region must be returned
or will flushed to the name be
truncated?

numlinks 31 uint32 READ
Number of links server
(although not necessarily to disk).

Note that flushing data to
this object.

owner 32 utf8<> R/W
The string name of the
owner of this object.

owner_group 33 utf8<> R/W
The string name server and the invalidation of cached
data must reflect the
group actual byte ranges locked or unlocked.
Rounding these up or down to reflect client cache block boundaries
will cause problems if not carefully done. For example, writing a
modified block when only half of that block is within an area being
unlocked may cause invalid modification to the owner region outside the
unlocked area which may be part of a region locked by another client.
Clients can avoid this object.

quota_hard 34 uint64 READ
Number of bytes situation by synchronously performing portions
of
disk space beyond write operations that overlap that portion (initial or final) that
is not a full block. Similarly, invalidating a locked area which is
not an integral number of full buffer blocks would require the server will
decline client
to allocate
new space.

quota_soft 35 uint64 READ
Number of bytes of
disk space at read one or two partial blocks from the server if the revalidation
procedure shows that the data which the client possesses may choose not be
valid.

Writes required to warn the user about
limited space.

quota_used 36 uint64 READ
Number of bytes of
disk space occupied flush data before unlocking must be done to stable
storage, either by doing synchronous writes or a COMMIT as part of
the owner flush operation. The is so because retransmission of this
object on this
filesystem.

rawdev 37 specdata4 READ
Raw device identifier.

Expires: December 1999 [Page 25]

Draft Protocol Specification NFS version 4 June 1999

space_avail 38 uint64 READ
Disk space the
modified data after a server reboot might conflict with a lock held
by another client.

Clients may choose to accommodate programs using record locking in bytes
available
non-standard ways (e.g. using a record lock as a global semaphore),
by flushing to this user
on the filesystem
containing this object
- this should be server more data upon an UNLOCK than is covered by
the
smallest relevant
limit.

space_free 39 uint64 READ
Free disk space locked range, possibly including modified data in
bytes on the
filesystem containing
this object - this
should be the smallest
relevant limit.

space_total 40 uint64 READ
Total disk space other files.
Any client doing so must ensure that for any file in
bytes on the
filesystem containing
this object.

space_used 41 uint64 READ
Number which all data
written is to properly locked areas, no piece of filesystem
bytes allocated data be written to
this object.

system 42 boolean R/W
Whether or
the server which is not this within the locked area.

9.4.3. Data Caching and Mandatory File Locking

Client side data caching needs to respect mandatory file locking when
this is a system file.

time_access 43 nfstime4 R/W in effect. The time presence of last
access to mandatory file locking for a
given file is indicated in the result flags for an OPEN. When there
is a read or write for a file for which mandatory locking is in
effect, the object.

time_backup 44 nfstime4 R/W
The time of last
backup of client must check if it holds an appropriate lock for the object.

time_create 45 nfstime4 R/W
The time of creation
range of bytes being read or written. If it does, it may satisfy the object. This
attribute does not
have
request using client side caching, just as for any relation to
the traditional Unix
file attribute 'ctime' other read or 'change time'.

time_delta 46 nfstime4 READ
Smallest useful server
time granularity.

time_metadata 47 nfstime4 R/W
The time of last
meta-data modification
of the object.

Expires: December 1999 [Page 26]

Draft Protocol Specification NFS version 4 June 1999

time_modify 48 nfstime4 R/W
The time since
write. If such a lock is not held, the
epoch of last
modification read or write cannot be
satisfied by caching but must be sent to the
object.

version 49 utf8<> R/W
Version number of this
document.

volatility 50 nfstime4 READ
Approximate time until
next expected change
on this filesystem, as server. When a measure of
volatility. request
partially overlaps a locked area, the request should be broken up
into multiple pieces with each region (locked or not) treated

Expires: December 1999 March 2000 [Page 27] 55]

Draft Protocol Specification NFS version 4 June September 1999

6. NFS Server Namespace

6.1. Server Exports

On a UNIX server the name-space describes all the files reachable by
pathnames under

appropriately.

9.4.4. Data Caching and File Identity

When clients cache data, data needs to organized according to the root directory "/". On a Windows NT server the
name-space constitutes all
file system object to which the files on disks named by mapped disk
letters. data belongs. For NFS server administrators rarely make v3 clients,
the entire server's
file-system name-space available typical practice has been to NFS clients. Typically, pieces
of the name-space are made available via an "export" feature. The
root filehandle for each export is obtained through the MOUNT
protocol; the client sends a string assume (for this purpose) that identifies
distinct handles represent distinct filesystem objects (even though
in some unusual cases this has not been the export of
name-space case) and the server returns the root filehandle for it. The
MOUNT protocol supports an EXPORTS procedure that will enumerate the
server's exports.

6.2. Browsing Exports

The data
cache may be maintained on the this basis.

In NFS version 4 protocol provides v4, we have the prospect (due to pathname based handles) of
more significant deviations from a root filehandle that one-filehandle-per-object model.
This requires some method by which clients
can use to obtain may reliably determine
whether two filehandles for these exports via a multi-component
LOOKUP. A common user experience is designate the same object. If they were to use a graphical user
interface (perhaps a file "Open" dialog window)
simply assume that all distinct filehandles denoted distinct objects
and proceeded to find a file via
progressive browsing through a directory tree. The do data caching on that basis, caching
inconsistencies would arise between the distinct client must side objects
which mapped to the same server side object. While it is true that
such inconsistencies would be
able similar to move those typically seen by
programs running on multiple clients (apart from one export to another export via single-component,
progressive LOOKUP operations.

This style of browsing is this issue), these
inconsistencies would not well supported by be expected an NFS version 2 and 3
protocols. The client expects all LOOKUP operations to remain within
a single server file-system, i.e. the device attribute will v3 clients not
change. This prevents sharing
files with any other client. The appearance of such inconsistencies
would be a client definite problem inhibiting transition from taking name-space paths that
span exports.

An automounter on the NFS v3 to NFS
v4 and so must be avoided.

The following procedure allows an NFS v4 client can obtain a snapshot of the server's
name-space using to determine (for the EXPORTS procedure
purposes of data caching) whether two distinct filehandles denote the MOUNT protocol.
same server side object:

o If it
understands GETATTR directed to the server's pathname syntax, it can create an image two handles in question have
different values of
the server's name-space fsid.major or fsid.minor, then they are
distinct objects.

o If GETATTR for any file on the client. The parts of fsid (major and minor) to which
the name-space
that are not exported by two handles belong and unique_handles is TRUE, then the server two
objects are filled in with a "pseudo
file-system" that allows the user distinct.

o If GETATTR directed to browse from the two handles does not return the
fileid attribute for one mounted file-
system to another. There is a drawback to this representation or both of the
server's name-space on handles, then the client: it is static. If
cannot be determined whether the server
administrator adds a new export two objects are the same and so
operations which depend on that knowledge (e.g. client will side data
caching) cannot be unaware of it. done reliably.

o If the two GETATTR's return different values for the fileid

Expires: December 1999 March 2000 [Page 28] 56]

Draft Protocol Specification NFS version 4 June September 1999

6.3. Server Pseudo File-System

NFS version 4 servers avoid this name-space inconsistency by
presenting all the exports within

attribute, then they are distinct objects.

o Otherwise they are the framework of same object.

9.5. Open Delegation

When a single server
name-space. An NFS version 4 client uses LOOKUP and READDIR
operations to browse seamlessly from one export to another. Portions
of file is being opened, the server name-space that are not exported are bridged via a
"pseudo file-system" that provides a view only may delegate further handling
of exported
directories. The pseudo file-system has a unique fsid and behaves
like a normal, read-only file-system.

6.4. Multiple Roots

DOS, Windows 95, 98 opens and NT are sometimes described as having
"multiple roots". File-Systems are commonly represented as disk
letters. MacOS represents file-systems as top-level names. NFS
version 4 servers closes for these platforms can construct a pseudo file-
system above these root names so that disk letters or volume names
are simply directory names in the pseudo-root.

6.5. Filehandle Volatility

The nature of file to the server's pseudo file-system opening client. Any such
delegation is recallable, since the circumstances that occasioned it is
are subject to change. In particular, the server may receive a logical
representation of file-system(s) available
conflicting OPEN from another client, which obliges it to recall the server.
Therefore, the pseudo file-system is most likely constructed
dynamically when
delegation before deciding whether the NFS version 4 is first instantiated. It OPEN may be granted. Granting
a delegation request is
expected up to the pseudo file-system server and it may deny all such
requests. The following is a typical set of conditions that servers
might use in deciding whether open should be delegated:

o The client must be able to respond to callbacks (as evidenced by
responding to previous CB_NULL requests).

o The client must not have an on-disk counterpart
from which persistent filehandles could failed to respond properly to previous
recalls.

o There must be no current open conflicting with the requested
delegation.

o There should be no current delegation that conflicts with the
delegation being requested.

o The probability of future conflicting open requests should be constructed. Even though
it is preferable that
low based on the recent history of the file.

o The existence of any server provide persistent filehandles for specific semantics of OPEN/CLOSE
that would make the pseudo file-system, required handling incompatible with the NFS client should expect
prescribed handling that pseudo
file-system file-handles are volatile. This can be confirmed by
checking the associated "persistent_fh" attribute for those
filehandles in question. If the filehandles delegated client would apply (see
below).

There are volatile, the NFS two types of open delegations, read and write. A read open
delegation allows a client must be prepared to recover a filehandle value (i.e. with a v4
multi-component LOOKUP) when receiving an error of NFS4ERR_FHEXPIRED.

6.6. Exported Root

If the server's root file-system is exported, it might be easy handle, on its own, requests to
conclude that open a pseudo-file-system is
file for reading that do not needed. This would deny read access to others. Multiple
read open delegations may be
wrong. Assume outstanding simultaneously and do not
conflict. A write open delegation allows the following file-systems client to handle on its
own all opens. Only one write open delegation may exist for a server:

/ disk1 (exported)
/a disk2 (not exported) given
file at a given time and it is inconsistent with any read open

Expires: December 1999 March 2000 [Page 29] 57]

Draft Protocol Specification NFS version 4 June September 1999

/a/b disk3 (exported)

Because disk2 is

delegations.

When a client has a read open delegation, it may not exported, disk3 cannot be reached with simple
LOOKUPs. The server must bridge make any changes
to the gap with contents or attributes of the file but it is assured that no
other client may do so. When a pseudo-file-system.

6.7. Mount Point Crossing

The server file-system environment client has a write open delegation it
may constructed modify the file data as it wishes secure in such a way the knowledge that
one file-system contains no
other client is accessing the file's data. The client holding a directory
write delegation may only affect file attributes which is 'covered' or mounted
upon by are intimately
connected with the file data: length, modify_time, change.

When a second file-system. For example:

/a/b (file system 1)
/a/b/c/d (file system 2)

The pseudo file-system for this client has an open delegation, it does not send OPEN's, or
CLOSE's to the server may but updates the appropriate status internally.
For a read open delegation, opens that cannot be constructed handled locally
(opens for write or that deny read access) must go to look
like:

/ (place holder/not exported)
/a/b (file system 1)
/a/b/c/d (file system 2)

It the server.

When an open delegation is requested and granted, the server's responsibility response to present the pseudo file-system
that is complete to
OPEN contains an open delegation structure which specifies, the client. If type
of delegation (read or write), space limitation information to
control flushing of data on close (write open delegation only, see
the client sends Section "Open Delegation and Data Caching"), an nfsacl4
specifying read and write permissions and a lookup request
for the path "/a/b/c/d", stateid to represent the server's response
delegation when doing IO. This stateid is separate and distinct from
the filehandle of stateid for the file system "/a/b/c/d". In previous versions of NFS, OPEN proper, which, unlike the server
would respond delegation
stateid, is associated with the directory "/a/b/d/d" within the file-system
"/a/b".

The NFS client a particular nfs_lockowner, and will be able
continue to determine be valid after the delegation is recalled, if it crosses the file
remains open.

When an internal request (or a server mount
point request by one of a change proxy server's
clients) is made to open a file when open delegation is in effect, it
will be accepted or rejected solely on the value basis of the "fsid" attribute.

6.8. Summary

NFS version 4 provides LOOKUP and READDIR operations following
conditions. Any requirement for browsing of
NFS file-systems. These operations are also used other checks to browse server
exports. A v4 server supports export browsing be made by including exported
directories the
delegate, should result in a pseudo-file-system. A browsing client open delegation being denied so that the
checks can cross
seamlessly between a pseudo-file-system be made by the server itself.

o The access and a real, exported file-
system. Clients must support volatile filehandles deny bits for the request and recognize
mount point crossing of server file-systems. the file as
described in Section 7.7, Share reservations

o The read and write permissions as determined below.

The nfsacl4 passed with delegation can be used to avoid frequent
ACCESS calls. The permission check should be as follows:

o If the nfsacl4 indicates that the open may be done, then it
should be granted, without reference to the server.

Expires: December 1999 March 2000 [Page 30] 58]

Draft Protocol Specification NFS version 4 June September 1999

7. File Locking

Integrating locking into NFS necessarily causes it to

o If the nfsacl4 indicates that the open may not be state-full,
with done, then an
ACCESS request must be made to the invasive nature of "share" file locks it becomes
substantially server to obtain the
definitive answer.

The server may thus return an nfsacl4 that is more dependent on state restrictive than
the traditional
combination actual ACL of NFS and NLM [XNFS]. There are three components the file, including one that specifies denial of
all access. Note that some common practices like mapping root to
making this state manageable:

o Clear division between client and server

o Ability
nobody may make it incorrect to reliably detect inconsistency send the actual ACL of the file in state between client
and server

o Simple and robust recovery mechanisms

In this model,
some cases.

The use of delegation together with various other forms of caching
creates the possibility that no server owns authentication will ever be
performed on a given user since all of his requests might be
satisfied locally. Where the state information. The client
communicates its view is depending of this state to the server as needed. for
authentication, it should make sure that some authentication (via an
ACCESS call) happens for each user, even if an ACCESS call would not
otherwise be required. The
client server may enforce frequent
authentication by returning an nfsacl4 denying all access with every
open delegation.

9.5.1. Open Delegation and Data Caching

Open delegation allows much of the message overhead associated with
opening and close files to be eliminated. This is also able to detect inconsistent state before modifying the case for
a
file.

To support Windows "share" locks, it is necessary proxy server to atomically which an open
or create files. Having a separate share/unshare operation will delegation was made but which did not
allow correct implementation of
pass the Windows OpenFile API. delegation on. In order
to correctly implement share semantics, the existing mechanisms used either case, an open when a file is opened or created (LOOKUP, CREATE, ACCESS) need to be
replaced. NFS V4 will have an OPEN procedure open
delegation was in effect would not require that subsumes a validation message
be sent to the
functionality server. The continued endurance of LOOKUP, CREATE, the read-open-
delegation provides a guarantee that no open for write and ACCESS. However, because many
operations require thus no
write has occurred. Similarly, when closing a file handle, the traditional LOOKUP opened for write,
if write open delegation is preserved in effect, the data written does not have
to map a file name be flushed to file handle without establishing state on the
server. Policy server until the open delegation is recalled.
The continued endurance of granting access the open delegation provides a guarantee
that no open and thus no read or modifying files is managed write has been done by another
client.

For the server based on purposes of open delegation, IO done without an OPEN (via
special stateid's consisting of all zero bits or all one bits) are
treated as the client's state. It is believed that these
mechanisms can implement policy ranging from advisory only locking to
full mandatory locking. While ACCESS is just functional equivalent of a subset corresponding type of open.
Thus, such READ's or WRITE's done by another client need will provoke
recall of OPEN, the
ACCESS procedure is maintained as a lighter weight mechanism.

7.1. Definitions

Lock The term "lock" write open delegation, will be used to refer any such WRITE will provoke
recall of a read open delegation.

In order to both record
(byte-range) locks as well as file (share) locks unless
specifically stated otherwise.

Client Throughout this proposal maintain current semantics in which the term "client" is used non-availability
of storage to
indicate the entity that maintains hold a set of locks on behalf
of one or more applications. The file written by an NFS client is responsible for
crash recovery of those locks it manages. Multiple clients
may share the same transport and multiple clients may exist guaranteed to

Expires: December 1999 March 2000 [Page 31] 59]

Draft Protocol Specification NFS version 4 June September 1999

be determined at or before the same network node.

Clientid A 64-bit quantity returned associated close operation, the
avoidance by a server that uniquely
corresponds to a the client supplied Verifier and ID.

Lease An interval of time defined by the requirement to flush data to the
server for on close, is limited to cases in which the client is irrevokeably granted a lock. At and server
together can determine in advance that the end required space will be
available. The server specifies one of a
lease period number of limiting
conditions, either a limit on the lock may be revoked if size of the lease has not
been extended. The lock must be revoked if file or a conflicting
lock has been granted after limit on the lease interval. All leases
granted by
number of modified blocks using a server have the same fixed interval.

Stateid A 64-bit quantity returned blocksize supplied by a server that uniquely
defines the locking state granted by server.
Based on implementation experience, changes in the server for a
specific lock owner for a specific file. A stateid
composed form of all bits 0 these
conditions may be made or all bits 1 have special meaning
and are reserved.

Verifier A 32-bit quantity generated by new types of limiting conditions defined.
Whatever the client that form of condition used, it us up to the server
can use to determine if the client has restarted and lost
all previous lock state.

7.2. Locking

It is assumed ensure
that manipulating any set of writes, no matter how arranged that meets the
specified condition will ever encounter a lock is rare lack of disk space
availability when compared the modified data is allowed to I/O
operations. It remain on the
client unflushed to the server past the point of close. The server
must make sure that the maximum possible amount of storage is also assumed
reserved so that crashes all outstanding delegations together meet that
condition, and network partitions
are relatively rare. Therefore to recall delegations appropriately to maintain that
invariant. When a server implements quotas, it should also be
careful that it does not invalidate its quota invariants when
granting write open delegation. When a user is important that I/O operations
have near a light weight mechanism quota limit,
this may result in write open delegations granted with very
restrictive space limitation conditions or those which always force
modified data to be flushed to indicate if they possess a held
lock. A lock request contains the heavy weight information required server on close.

When authentication considerations make flushing of modified data to establish a lock and uniquely define
the lock owner.

The following sections describe server after the transition from close problematic (after the heavy weight
information to last close, the eventual stateid used for most client and server
locking
user may have logged off and lease interactions.

7.2.1. Client ID

For each LOCK request, unexpired local credentials may not
exist), the client must identify itself may need to take special care to ensure that local
unexpired credentials will in fact be available, either by tracking
the server.
This is done expiration time of credentials and flushing data well in such a way as advance
of their expiration, or by making private copies of credentials to allow for correct lock
identification
assure their availability when needed.

9.5.2. Open Delegation and crash recovery. Client identification is
accomplished with two values.

o A verifier that is used to detect File Locks

When a client reboots.

o A variable length opaque array to uniquely define holds a client.

For an operating system this may write-open-delegation, lock operations,
including those required by mandatory file locking are performed
locally since the delegation implies that there can be no conflicting
locks. On a fully qualified host similar basis, all of the revalidations that would
normally be associated with obtaining locks and the flushing of data
which would attend the releasing of locks for write need not be done.

9.5.3. Recall of Open Delegation

Expires: December 1999 March 2000 [Page 32] 60]

Draft Protocol Specification NFS version 4 June September 1999

name or IP address, and for a user level NFS client it may
additionally contain a process id or other unique sequence.

The data structure for the Client ID would then appear as:
struct nfs_client_id {
opaque verifier[4];
opaque id<>;
}:

It is possible through the mis-configuration of a client or the
existence of a rogue client that two clients end up using the same
nfs_client_id. This situation is avoided by 'negotiating' the
nfs_client_id between client and server with the use of the
SETCLIENTID.

The following describes the two scenarios events necessitate recall of
negotiation.

1 Client has never connected to the server

In this case the client generates an nfs_client_id and
unless open delegation:

o Potentially conflicting OPEN request (or IO done with "special"
stateid)

o SETATTR issued by another client has the same nfs_client_id.id field,
the server accepts the request. The server also records the
principal (or principal to uid mapping) from

o REMOVE request for the credential file in the RPC question

o RENAME request that contains for the nfs_client_id
negotiation request.

Two clients might still use file in question as either source or
target of the same nfs_client_id.id due
to perhaps configuration error (say a High Availability
configuration where RENAME

NOTE: [[The following are necessary unless the nfs_client_id.id spec is derived from
the ethernet controller address
cleaned up to disallow LOCK's and both systems have the
same address). In this case, nfs4err can be IO operations without a switched
union that returns
corresponding OPEN.]]

o LOCK request by another client.

o IO operation done with "special" stateid by another client.

Whether a RENAME of a directory in addition the path leading to NFS4ERR_CLID_IN_USE, the
network address (the rpcbind netid and universal address) file
results in recall of an open delegation depends on the semantics of
the client server file system. If that filesystem denies such RENAME's when
a file is using open, the id.

2 Client is re-connecting recall must be performed to determine whether the
file in question is, in fact, open.

In addition to the situations above, the server after may choose to recall
open delegations at any time if resource constraints make it
advisable to do so. Clients should always be prepared for the
possibility of recall.

Special handling is needed for a client reboot GETATTR which occurs when a write
open delegation is in effect. In this case, the client still generates an nfs_client_id
but holding the nfs_client_id.id field will
delegation needs to be interrogated, using a CB_GETATTR callback, if
the same as the
nfs_client_id.id generated prior to reboot. If GETATTR attribute bits include any of the server
finds attributes that the principal/uid is equal a write
open delegate may modify (length, modify time, change).

When a client receives a recall for an open delegation, it needs to
update state on the previously
"registered" nfs_client_id.id, then locks associated with
the old nfs_client_id are immediately released. If server before returning the
principal/uid is not equal, then this is delegation. These
same updates must be done whenever a rogue client and
the request is returned in error. For more discussion chooses to return a
delegation voluntarily. The following items of
crash recovery semantics, see the section on "Crash
Recovery" state need to be
dealt with:

Expires: December 1999 March 2000 [Page 33] 61]

Draft Protocol Specification NFS version 4 June September 1999

In both cases, upon success, NFS4_OK is returned. To help reduce

o If the open file associated with the
amount of data transferred on OPEN and LOCK, which delivered the server will also
return a unique 64-bit clientid value that
delegation to the client is no longer open, then a short hand reference CLOSE must be
done to the nfs_client_id values presented by the client. From server, if this point
forward, has not been done previously.

o If there are other opens extant for that file, then OPEN
operations must be done to update the client can use server and obtain the clientid
stateid's to refer to itself.

7.2.2. nfs_lockowner and be used subsequently, given that the delegation
stateid definition

When requesting will no longer be valid. Such OPEN's are done using a lock,
claim type of CLAIM_DELEGATE_CUR so that the client must present delegation stateid
can be presented to the server to establish the
clientid and an identifier for the owner of client's right
to perform this OPEN. (See the requested lock.
These two fields section "Procedure 17: OPEN" for
details).

o If there are referred locks which have been granted (write open
delegation case only), then these need to be performed to as the nfs_lockowner and
server.

In the
definition case of those fields are:

o A clientid returned by a write open delegation, if the server as part of file in question is
not opened for write at the clients use time of recall, then any modified data
for the SETCLIENTID procedure

o A variable length opaque array used file needs to uniquely define the owner
of a lock managed by the client.

This may be a thread id, process id, or other unique value.

When the server grants the lock it responds with a unique 64-bit
stateid. The stateid is used as a short hand reference flushed to the
nfs_lockowner, since server, as it would have been
flushed when the server will be maintaining file was closed, had the
correspondence between them.

7.2.3. Use write open delegation not
been in effect. The possibility of truncation on the stateid

All I/O requests contain a stateid. If client means
that the nfs_lockowner performs
I/O on a range of bytes within following needs to be done:

o If a locked range, file truncate has been done on the stateid returned
by client (as part of an
OPEN UNCHECKED, for example), and this has not yet been
propagated to the server (it must be used before allowing any new
data to be written to indicate the appropriate lock (record
or share) is held. If no state is established by the client, either
record lock or share lock, a stateid server), it must be done as part of all bits 0 is used. If no
conflicting locks are held on the file, the server may grant
recall, again before writing modified data to the I/O
request. If a conflict with an explicit lock occurs, server.

o Any modified data for the request is
failed (NFS4ERR_LOCKED). This allows "mandatory locking" file needs to be
implemented.

A stateid of all bits 1 allows read requests flushed to bypass locking checks
at the the
server. However,

In the case of write requests with stateid with bits all 1
does not bypass open delegation, file locking imposes some
additional requirements.

An explicit The flushing of any modified data in any
area for which a write lock may not be granted was released while an I/O operation with
conflicting implicit the write open
delegation was in effect is what is required to precisely maintain
the associated invariant. However, because the write open delegation
implies no other locking by other clients, a simpler implementation
is being performed. to flush all modified data for the file (as described just above)
if any write lock has been released while the write open delegation
was in effect.

Expires: December 1999 March 2000 [Page 34] 62]

Draft Protocol Specification NFS version 4 June September 1999

The byte range of

9.5.4. Delegation Revocation

When a lock delegation is indivisible. A range may revoked, if there are associated opens on the
client, the processes holding these opens need to be locked,
unlocked, notified,
normally by returning errors whenever IO operations or changed between read and write but a close is
attempted on that open file.

When an open delegation is revoked, if no opens are present on the
client, then no error needs to be reported, unless there is modified
data present on the client. In this case, the user will have to be
notified, since there may not have
subranges unlocked or changed between read be an active application to get an
error status. (See the section "Revocation Recovery for Write Open
Delegation" for more details).

9.6. Data Caching and write. This is Revocation

When locks (including delegations) are revoked, the
semantics provided by Win32 but only a subset assumptions upon
which successful caching depend, are no longer guaranteed. Therefore
the client, in addition to notifying the owner of a record lock or
share reservation, and processes holding opens for the semantics
provided by Unix. It is expected that Unix clients can more easily
simulate modifying subranges than Win32 servers adding this feature.

7.2.4. Sequencing delegation,
needs to remove all data for the file from its cache. In the case of
modified data, it must be removed from the client's cache without
being written to the server.

Notification to the lock requests

Locking is different than most NFS owner will in many cases consist of simply
returning an error on the next (and all subsequent) IO to the open
file or on the close. Where the client API make such notification
impossible (because errors for certain operations may not be
returned), more drastic action such as it requires "at-
most-one" semantics signals or process termination
may be appropriate since an invariant that an application depends on
may be violated. Depending on how errors are not provided by ONC RPC. In typically treated on
the face client operating system, further levels of
retransmission or reordering, lock or unlock requests must have a
well defined notification including
logging, console messages, and consistent behavior. To accomplish this each lock
request contains GUI pop-up's may be in order.

9.6.1. Revocation Recovery for Write Open Delegation

Revocation recovery for a sequence number write open delegation poses the issue in
that there may be modified data in the client cache while the file is a monotonically increasing
integer. Different nfs_lockowners have different sequences. The
not open. In this situation, any client which does not flush
modified data to the server maintains on each close must make sure that the last sequence number (L) received and
user receives appropriate notification of the
response that was returned. failure. Since such
situations may require human action to correct problems, notification
schemes in which the appropriate user or administrator is notified

Expires: March 2000 [Page 63]

Draft Protocol Specification NFS version 4 September 1999

may be necessary. Logging and console messages are typical examples.

If a request with a previous sequence
number (r < L) there is received modified data on the client, it is silently ignored as its response must have been received before not be flushed
normally to the last request (L) was sent. If server. A client may attempt to provide a
duplicate copy of last request (r == L) is received,
the stored response
is returned. If file data as modified during the delegation under a request beyond different
name, to ease recovery. Unless the next sequence (r == L + 2) is
received it is silently ignored. Sequences are reinitialized
whenever client can determine that the
file was has not modified by any other client, this technique is
limited to situations in which a client verifier changes.

7.3. Blocking locks

Some clients require has a complete cached copy of
the support file in question. Use of blocking locks. The current
proposal lacks such a call-back mechanism, similar to NLM, technique may be limited to notify
files under a
client certain size or may only be used when sufficient disk
space is guaranteed available within the target file system and when
the lock client has been granted. Clients have no choice but sufficient buffering resources to
continually poll for keep the lock, which presents a fairness problem.
Two new lock types cached copy
available until it is properly stored to the target file system.

9.7. Attribute Caching

First note that when attributes are added, READW discussed here, extended or named
attributes are not included. Individual named attributes are
analogous to files and WRITEW used caching of the data for these needs to indicate be
handled just as data caching is for ordinary files. Similarly, LOOKUP
results from an OPENATTR directory are to be cached on the same basis
as any other pathnames and similarly for directory contents.

Clients may cache file attributes obtained from the server and use
them to avoid subsequent GETATTR requests. Such caching is write
through in that the client modification to file attributes is requesting a blocking lock. The always done by
means of requests to the server and should maintain an ordered list of pending blocking locks. When not be done locally and
cached, the
conflicting lock is released, exception being modifications to attributes that are
intimately connected with data caching. Thus, extending a file by
writing data to the server may wait local data cache is reflected immediately in the lease period
for
length as seen on the first client to re-request without this change being immediately
reflected on the lock. After server. Normally such changes are not propagated
directly to the lease period
expires server, but when the next waiting client request modified data is allowed flushed to the lock. Clients
server, analogous attribute changes are required to poll at an interval sufficiently small that it made on the server. When
open delegation is
likely in effect, the modified attributes may be returned
to acquire the lock server in the response to a timely manner. CB_RECALL call.

The server result of local caching of attributes is that the attribute
caches maintained on individual clients will not
required to maintain be coherent. Changes
made in one order on the server may be seen in a list of pending blocked locks as it is used to
increase fairness different order on
one client and in a third order on a different client.

Given that typical file API's do not correct operation. Because of provide means to atomically
modify or interrogate attributes for multiple files at the
unordered nature of crash recovery, storing same time,
the undesirable effects of lock state to stable
storage would be required to guarantee ordered granting these incoherencies have proved
manageable, if the following rules, derived from the practice of blocking
locks.

Expires: December 1999 March 2000 [Page 35] 64]

Draft Protocol Specification NFS version 4 June September 1999

7.4. Lease renewal

The purpose of a lease is to allow a server to remove stale locks
that

NFSv3 implementations are held by a client that has crashed or is otherwise
unreachable. It is not a mechanism followed:

o All attributes for cache consistency and lease
renewals may not be denied if the lease interval has not expired.
Any I/O request that has been made with a valid stateid is given file (per-fsid attributes excepted)
are cached as a positive
indication unit so that the client is still alive and locks are being
maintained. This becomes an implicit renewal of the lease. In the
case no I/O has been performed non-serializability can arise
within the lease interval, context of a lease single file.

o A bound is maintained on how long a client cache entry can be renewed by having the client issue a zero length READ. Because
kept without being refreshed from the nfs_lockowner contains a unique client value, server.

o When performing any stateid for a
client will renew all leases for locks held with the same client
field. This will allow very low overhead lease renewal that scales
extremely well. In the typical case, no extra RPC calls are needed
and in the worst case one RPC is required every lease period
regardless of the number of locks held by the client.

7.5. Crash recovery

The important requirement in crash recovery is operation that both the client
and the server know when changes attributes on the other has failed. Additionally it is
required that a client sees a consistent view of data across server
reboots. I/O
server, including directory operations that may have been queued within the client
or network buffers, cannot complete until after the client has
successfully recovered the lock protecting due so indirectly,
updated attributes would be fetched as part of the I/O operation.

If associated
rpc, using a client fails, the server only needs to wait the lease period to
allow conflicting locks. If the client reinitializes within GETATTR following the
lease period, it may be forced to wait operation in question, which
the remainder results of the period
before resuming service. To minimize this delay, lock requests
contain a verifier field in GETATTR used to update the lock_owner, client's attribute
cache.

Note that if the server receives a
verifier field that does not match full set of attributes to be cached is requested by
READDIR, the existing verifier, results can be cached by the server
knows that client on the same basis as
attributes obtained GETATTR.

A client has lost all lock state and locks held may validate its cached version of attributes for a file by
fetching only the change attribute and assuming that
client that do not match if the current verifier may be released. In a
secure environment, a change in
attribute has the verifier must only cause same value as it did when the
locks held by attributes were
cached, then no attributes have changed, with the authenticated requester to possible exception
of access_time.

9.8. Name Caching

The results of LOOKUP and READDIR operations may be released in order cached to
prevent a rogue user from freeing otherwise valid locks. The
verifier must have avoid
the same uniqueness properties cost of subsequent LOOKUP operations. Just as in the COMMIT
verifier.

If the server fails and loses locking state, case
attribute caching, inconsistencies may arise among the server must wait various client
caches. To mitigate the
lease period before granting any new locks or allowing any I/O. An
I/O request during effects of these inconsistencies, given the grace period with an invalid stateid will fail
with NFS4ERR_GRACE,
context of typical file API's, the client will reissue following rules should be adhered
to:

o The results of unsuccessful LOOKUP's should not cached, unless
they are specifically reverified at the lock request with
reclaim set to TRUE, and upon receiving point of use.

o A bound is maintained on how long a successful reply, the I/O
may client name cache entry can
be reissued with kept without verifying that the new stateid. Any time entry in question has not
been made invalid by a client receives an directory change operation performed by
another client.

Expires: December 1999 March 2000 [Page 36] 65]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_GRACE error it should start recovering all outstanding locks.
A lock request during the grace period without reclaim set will also
result in

When a NFS4ERR_GRACE, triggering the client recovery processing.
A lock request outside the grace period with reclaim set will succeed
only if the server can guarantee that no conflicting lock or I/O
request has been granted since reboot.

In the case of a network partition longer than the lease period, the
server will have is not received an implicit lease renewal and may free
all locks held making changes to a directory for the client, thus invalidating any stateid held by
the client. Subsequent reconnection will cause I/O with invalid
stateid which there
exist name cache entries, it needs to fail with NFS4ERR_EXPIRED, the client will suitably notify
the application holding the lock. periodically fetch attributes
for that directory to make sure that it is not changing. After the lease period
determining that no change that has expired occurred, the server expiration time for
the associated name cache entries may optionally continue be updated to hold be the locks for current
time plus the client.
In this case, if name cache staleness bound.

When a conflicting lock or I/O request client is received, the
lock must be freed making changes to allow the client a given directory, it needs to detect possible corruption.
When
determine whether there is a network partition have been changes made to the directory by
other clients. It does this using the change attribute as reported
before and after the lease expires, directory operation in the server
must record associated wcc4_info
returned on stable storage that operation. When the client information relating to
those leases. This server is able to report these
values atomically with respect to prevent the case where another client
obtains directory operation, which the conflicting lock, frees
server indicates in the lock, and wcc4_info, comparison of the server reboots.
After pre-operation
change value with the server recovers change value which the original client may recover has in his cache
determines whether there has been a change by another client,
necessitating a purge of name cache associated with the network
partition and attempt to reclaim directory.
If there has been no such change, the lock. Without any state to
indicate that a conflicting may have occurred, name cache can be updated on
the client could get
in an inconsistent state. Storing just to reflect the client information is directory operation and the
minimal state necessary to detect this condition, but could lead to
losing locks unnecessarily. However this is considered associated
timeout extended. The post-operation change value needs to be a very
rare event, and a sophisticated server could store more state
completely eliminate any unnecessary locks being lost.

7.6. Server revocation of locks

The server can revoke saved
as the locks held basis for future wcc4_info comparisons.

Name caching requires that the client revalidate cached data by
comparing the change attribute for a client at any time, directory when the client detects revocation it must ensure its state matches name item was
cached. This requires that
of the server. If locks are revoked due to a server reboot, any changes in the
client will receive contents of a NFS4ERR_GRACE and normal crash recovery
described above will
directory be performed.

The server may revoke visible as a lock within changed value for the lease period, this is
considered change attribute of
the directory. Proper use of wcc4_info, when a rare event likely client makes a change
to be initiated only by a human (as
part of an administration task). The client may assume directory, requires that only reporting of the
file that caused pre-operation and
post-operation change attribute values are in fact atomic with the NFS4ERR_EXPIRED
actual directory change. When the server cannot reliably report
before and after values atomically with respect to be returned has lost the
lock_owner's locks directory
operation, the server indicates that in the wcc4_info and notifies the holder appropriately. The client
can
should not assume that other clients have not changed the lease period has been renewed. directory.

9.9. Directory Caching

The client not being able results of READDIR operations may be used to renew avoid subsequent
READDIR operations. Just as in the lease period is a relatively
rare cases of attribute and unusual state. Both sides will detect name
caching, this state and can
recover without data corruption. The may result in inconsistencies among the various client
caches. To mitigate the effects of these inconsistencies, given the
context of typical file API's, the following rules should be adhered
to:

o Cached READDIR information for a directory which is not obtained
in a single READDIR operation must mark all locks held always be a consistent

Expires: December 1999 March 2000 [Page 37] 66]

Draft Protocol Specification NFS version 4 June September 1999

as "invalidated" and then must issue an I/O request, either a pending
I/O or zero length read to revalidate the lock. If the response is
success the lock is upgraded to valid, otherwise it was revoked by
the server and the owner is notified.

7.7. Share Reservations

A share reservation is a mechanism to control access to a file. It
is a separate and independent mechanism from record locking. When a
client opens a file, it issues an OPEN request to the server
specifying the type

snapshot of access required (READ, WRITE, or BOTH) directory contents as evidenced by a GETATTR before
the first and after the
type last of access to deny others (deny NONE, READ, WRITE, or BOTH). If READDIR's which contribute.

o A bound is maintained on the OPEN fails amount of time that a directory
cache entry may be kept on the client will fail without revalidation.

The revalidation technique parallels that discussed in the applications open request.

Pseudo-code definition case of
name caching. When the semantics:

if ((request.access & file_state.deny)) ||
(request.deny & file_state.access))
return (NFS4ERR_DENIED)

Old DOS applications specify shares in compatibility mode. Microsoft
has indicated in client is not changing the Win32 specification that it will be deprecated directory in the future and recommends
question, checking that deny NONE be used. This
specification does the directory has not support compatibility mode.

7.8. OPEN/CLOSE procedures

To provide correct share semantics, changed (by using
GETATTR to obtain the change attribute) is adequate to extend the
lifetime of the cache entry. When a client MUST is modifying the
directory, it needs to use the OPEN
procedure wcc4_info data to obtain determine whether
there are other clients who are modifying the initial file handle and indicate directory, allowing it
to update the desired
access and what if any access directory cache to deny. Even reflect its own changes if it is the
only client intends to
use a stateid of all 0's or all 1's, it must still obtain making modifications.

Directory caching requires that the
filehandle for client revalidate cached data by
comparing the regular file with change attribute for a directory when the OPEN procedure. For clients directory
data was cached. This requires that do not have any changes in the contents of a deny mode built into their open API, deny equal to
NONE should
directory be used.

The OPEN procedure with visible as a changed value for the CREATE flag, also subsumes change attribute of
the CREATE
procedure for regular files as used in previous versions directory. Proper use of NFS,
allowing wcc4_info, when a create with client makes a share change
to be done atomicly.

Will expand on create semantics here.

The CLOSE procedure removes all share locks held by the lock_owner on
that file. If record locks are held they should be explicitly

Expires: December 1999 [Page 38]

Draft Protocol Specification NFS version 4 June 1999

unlocked. Some servers may not support the CLOSE of a file directory, require that
still has record locks held; if so, CLOSE will fail reporting of the pre-operation and return an
error.

The LOOKUP procedure is preserved
post-operation change attribute values are in fact atomic with the
actual directory change. When the server cannot reliably report
before and will return a file handle
without establishing any lock state on after values atomically with respect to the server. Without a valid
stateid, directory
operation, the server will assume indicates that in the wcc4_info and the client has
should not assume that other clients have not changed the least access. For
example, a file opened with deny READ/WRITE cannot be accessed using
a file handle obtained through LOOKUP. directory.

Expires: December 1999 March 2000 [Page 39] 67]

Draft Protocol Specification NFS version 4 June September 1999

10. Defined Error Numbers

NFS error numbers are assigned to failed operations within a compound
request. A compound request contains a number of NFS operations that
have their results encoded in sequence in a compound reply. The
results of successful operations will consist of an NFS4_OK status
followed by the encoded results of the operation. If an NFS
operation fails, an error status will be entered in the reply and the
compound request will be terminated.

A description of each defined error follows:

NFS4_OK Indicates the operation completed successfully.

NFS4ERR_PERM Not owner. The operation was not allowed
because the caller is either not a privileged
user (root) or not the owner of the target of
the operation.

NFS4ERR_NOENT No such file or directory. The file or
directory name specified does not exist.

NFS4ERR_IO I/O error. A hard error (for example, a disk
error) occurred while processing the requested
operation.

NFS4ERR_NXIO I/O error. No such device or address.

NFS4ERR_ACCES Permission denied. The caller does not have the
correct permission to perform the requested
operation. Contrast this with NFS4ERR_PERM,
which restricts itself to owner or privileged
user permission failures.

NFS4ERR_EXIST File exists. The file specified already exists.

NFS4ERR_XDEV Attempt to do a cross-device hard link.

NFS4ERR_NODEV No such device.

Expires: December 1999 March 2000 [Page 40] 68]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_NOTDIR Not a directory. The caller specified a non-
directory in a directory operation.

NFS4ERR_ISDIR Is a directory. The caller specified a
directory in a non-directory operation.

NFS4ERR_INVAL Invalid argument or unsupported argument for an
operation. Two examples are attempting a
READLINK on an object other than a symbolic
link or attempting to SETATTR a time field on a
server that does not support this operation.

NFS4ERR_FBIG File too large. The operation would have caused
a file to grow beyond the server's limit.

NFS4ERR_NOSPC No space left on device. The operation would
have caused the server's file system to exceed
its limit.

NFS4ERR_ROFS Read-only file system. A modifying operation
was attempted on a read-only file system.

NFS4ERR_MLINK Too many hard links.

NFS4ERR_NAMETOOLONG The filename in an operation was too long.

NFS4ERR_NOTEMPTY An attempt was made to remove a directory that
was not empty.

NFS4ERR_DQUOT Resource (quota) hard limit exceeded. The
user's resource limit on the server has been
exceeded.

NFS4ERR_STALE Invalid file handle. The file handle given in
the arguments was invalid. The file referred to
by that file handle no longer exists or access
to it has been revoked.

Expires: December 1999 March 2000 [Page 41] 69]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_BADHANDLE Illegal NFS file handle. The file handle failed
internal consistency checks.

NFS4ERR_NOT_SYNC Update synchronization mismatch was detected
during a SETATTR operation.

NFS4ERR_BAD_COOKIE READDIR cookie is stale.

NFS4ERR_NOTSUPP Operation is not supported.

NFS4ERR_TOOSMALL Buffer or request is too small.

NFS4ERR_SERVERFAULT An error occurred on the server which does not
map to any of the legal NFS version 4 protocol
error values. The client should translate this
into an appropriate error. UNIX clients may
choose to translate this to EIO.

NFS4ERR_BADTYPE An attempt was made to create an object of a
type not supported by the server.

NFS4ERR_JUKEBOX The server initiated the request, but was not
able to complete it in a timely fashion. The
client should wait and then try the request
with a new RPC transaction ID. For example,
this error should be returned from a server
that supports hierarchical storage and receives
a request to process a file that has been
migrated. In this case, the server should start
the immigration process and respond to client
with this error.

NFS4ERR_SAME Returned if an NVERIFY operation shows that no
attributes have changed.

NFS4ERR_DENIED An attempt to lock a file is denied. Since
this may be a temporary condition, the client
is encouraged to retry the lock request (with
exponential backoff of timeout) until the lock

Expires: December 1999 March 2000 [Page 42] 70]

Draft Protocol Specification NFS version 4 June September 1999

is accepted.

NFS4ERR_EXPIRED A lease has expired that is being used in the
current procedure.

NFS4ERR_LOCKED A read or write operation was attempted on a
locked file.

NFS4ERR_GRACE The server is in its recovery or grace period
which should match the lease period of the
server.

NFS4ERR_FHEXPIRED The file handle provided is volatile and has
expired at the server. The client should
attempt to recover the new file handle by
traversing the server's file system name space.
The file handle may have expired because the
server has restarted, the file system object
has been removed, or the file handle has been
flushed from the server's internal mappings.

NOTE: This error definition will need to be crisp and match
the section describing the volatile file handles.

NFS4ERR_SHARE_DENIED An attempt to OPEN a file with a share
reservation has failed because of a share
conflict.

NFS4ERR_SAME This error is returned by the NVERIFY operation
to signify that the attributes compared were
the same as provided in the client's request.

NFS4ERR_WRONGSEC The security mechanism being used by the client
for the procedure does not match the server's
security policy. The client should change the
security mechanism being used and retry the
operation.

NFS4ERR_CLID_INUSE The SETCLIENTID procedure has found that a

Expires: December 1999 March 2000 [Page 43] 71]

Draft Protocol Specification NFS version 4 June September 1999

client id is already in use by another client.

NFS4ERR_RESOURCE For the processing of the COMPOUND procedure,
the server may exhaust available resources and
can not continue processing procedures within
the COMPOUND operation. This error will be
returned from the server in those instances of
resource exhaustion related to the processing processing
of the COMPOUND procedure.

NFS4ERR_MOVED The filesystem which contains the current
filehandle object has been relocated or
migrated to another server. The client may
obtain the new filesystem location by obtaining
the "fs_locations" attribute for the current
filehandle. For further discussion, refer to
the section "Filesystem Migration or
Relocation".

NFS4ERR_NOFILEHANDLE The logical current file handle value has not
been set properly. This may be a result of the a
malformed COMPOUND procedure. operation (i.e. no PUTFH or
PUTROOTFH before an operation that requires the
current file handle be set).

Expires: December 1999 March 2000 [Page 44] 72]

Draft Protocol Specification NFS version 4 June September 1999

11. NFS Version 4 Requests

For the NFS program, version 4, there are two traditional RPC
procedures: NULL and COMPOUND. All other procedures operations for NFS version
4 are defined in normal XDR/RPC syntax and semantics except that
these procedures operations are encapsulated within the COMPOUND request. This
requires that the client combine one or more NFSv4 procedures operations into a
single request.

9.1.

The NFS4_CALLBACK program is used to provide server to client
signaling and is constructed in a similar fashion as the NFS program.
The procedures CB_NULL and CB_COMPOUND are defined in the same way as
NULL and COMPOUND are within the NFS program. The CB_COMPOUND
request also encapsulates the remaining operations of the
NFS4_CALLBACK program.

11.1. Compound Procedure

These compound requests provide the opportunity for better
performance on high latency networks. The client can avoid
cumulative latency of multiple RPCs by combining multiple dependent
operations into a single compound request. A compound op may provide
for protocol simplification by allowing the client to combine basic
procedures into a single request that is customized for the client's
environment.

The basics of the COMPOUND procedures construction is:

+-----------+-----------+-----------+--
| op + args | op + args | op + args |
+-----------+-----------+-----------+--

and the reply looks like this:

+----------------+----------------+----------------+--
| code + results | code + results | code + results |
+----------------+----------------+----------------+--

Where "code" is an indication of the success or failure of the
operation including the opcode itself.

9.2.

11.2. Evaluation of a Compound Request

The server will process the COMPOUND procedure by evaluating each of
the procedures operations within the COMPOUND request in order. Each component
procedure or

Expires: March 2000 [Page 73]

Draft Protocol Specification NFS version 4 September 1999

operation consists of a 32 bit operation code, followed by the
argument of length determined by the type of operation. The results
of each operation are encoded in sequence into a reply buffer. The
results of each operation are preceded by the opcode and a status
code (normally zero). If an operation results in a non-zero status
code, the status will be encoded and evaluation of the compound
sequence will halt and the reply will be returned.

Expires: December 1999 [Page 45]

Draft Protocol Specification NFS version 4 June 1999

There are no atomicity requirements for the procedures contained
within the COMPOUND procedure. The procedures operations being evaluated as
part of a COMPOUND request may and more than likely will be evaluated simultaneously with other
COMPOUND requests that the server receives.

It is the client's responsibility for recovering from any partially
completed compound request.

Each operation assumes a "current" filehandle that is available as
part of the execution context of the compound request. Operations
may set, change, or return this filehandle.

Expires: December 1999 March 2000 [Page 46] 74]

Draft Protocol Specification NFS version 4 June September 1999

10.

12. NFS Version 4 Procedures

10.1.

12.1. Procedure 0: NULL - No Operation

SYNOPSIS

<null>

ARGUMENT

void;

RESULT

void;

DESCRIPTION

Standard ONCRPC NULL procedure. Void argument, void response.

ERRORS

None.

Expires: December 1999 March 2000 [Page 47] 75]

Draft Protocol Specification NFS version 4 June September 1999

10.2.

12.2. Procedure 1: COMPOUND - Compound Operations

SYNOPSIS

compoundargs -> compoundres

ARGUMENT

union opunion switch (unsigned opcode) {
case <OPCODE>: <argument>;
...
};

struct op {
opunion ops;
};

struct COMPOUND4args {
utf8string tag;
op oplist<>;
};

RESULT

struct COMPOUND4resok COMPOUND4res {
nfsstat4 status;
utf8string tag;
resultdata data<>;
};

union COMPOUND4res switch (nfsstat4 status){
case NFS4_OK:
COMPOUND4resok resok4;
default:
void;
};

DESCRIPTION

The COMPOUND procedure is used to combine one or more of the NFS
procedures into a single RPC request. The main NFS RPC program has
two main procedures: NULL and COMPOUND. All other procedures use
the COMPOUND procedure as a wrapper.

Expires: December 1999 [Page 48]

Draft Protocol Specification NFS version 4 June 1999

In the processing of the COMPOUND procedure, the server may find
that it does not have the available resources to execute any or all
of the procedures within the COMPOUND sequence. In this case, the
error NFS4ERR_RESOURCE will be returned for the particular
procedure within the COMPOUND operation where the resource
exhaustion occurred. This assume that all previous procedures

Expires: March 2000 [Page 76]

Draft Protocol Specification NFS version 4 September 1999

within the COMPOUND sequence have been evaluated successfully.

IMPLEMENTATION

The COMPOUND procedure is used to combine individual procedures
into a single RPC request. The server interprets each of the
procedures in turn. If a procedure is executed by the server and
the status of that procedure is NFS4_OK, then the next procedure in
the COMPOUND procedure is executed. The server continues this
process until there are no more procedures to be executed or one of
the procedures has a status value other than NFS4_OK.

Note that the definition of the "tag" in both the request and
response are left to the implementor. It may be used to summarize
the content of the compound request for the benefit of packet
sniffers and engineers debugging implementations.

ERRORS

NFS4ERR_RESOURCE

Expires: December 1999 March 2000 [Page 49] 77]

Draft Protocol Specification NFS version 4 June September 1999

10.3. Procedure

12.2.1. Operation 2: ACCESS - Check Access Permission Rights

SYNOPSIS

(cfh), permbits accessreq -> permbits supported, accessrights

ARGUMENT

const ACCESS4_READ= ACCESS4_READ = 0x0001;
const ACCESS4_LOOKUP= ACCESS4_LOOKUP = 0x0002;
const ACCESS4_MODIFY= ACCESS4_MODIFY = 0x0004;
const ACCESS4_EXTEND= ACCESS4_EXTEND = 0x0008;
const ACCESS4_DELETE= ACCESS4_DELETE = 0x0010;
const ACCESS4_EXECUTE= ACCESS4_EXECUTE = 0x0020;

struct ACCESS4args {
/* CURRENT_FH: object */
uint32_taccess;
uint32_t access;
};

RESULT

struct ACCESS4resok {
uint32_taccess;
uint32_t supported;
uint32_t access;
};

union ACCESS4res switch (nfsstat4 status) {
case NFS4_OK:
ACCESS4resokresok;
ACCESS4resok resok;
default:
void;
};

DESCRIPTION

ACCESS determines the access rights that a user, as identified by
the credentials in the request, has with respect to a file system
object. The client encodes the set of permissions access rights that are to be
checked in a bit mask. The server checks the permissions encoded
in the bit mask. A If a status of NFS4_OK is returned along with a returned, two bit
mask encoded with the permissions that the client is allowed. masks

Expires: December 1999 March 2000 [Page 50] 78]

Draft Protocol Specification NFS version 4 June September 1999

are included in the response. The first represents the access
rights for which the server can verify reliably for the user. The
second represents the access rights available to the user for the
filehandle provided.

The results of this procedure are necessarily advisory in nature.
That is, a return status of NFS4_OK and the appropriate bit set in
the bit mask does not imply that such access will be allowed to the
file system object in the future, as access rights can be revoked
by the server at any time.

The following access permissions may be requested:

ACCESS_READ: bit 1 Read data from file or read
a directory.
ACCESS_MODIFY: bit 2 Rewrite existing file data or modify
existing directory entries.
ACCESS_LOOKUP: bit 3 2 Look up a name in a
directory (no meaning for
non-directory objects).
ACCESS_MODIFY: bit 3 Rewrite existing file data or modify
existing directory entries.
ACCESS_EXTEND: bit 4 Write new data or add
directory entries.
ACCESS_DELETE: bit 5 Delete an existing
directory entry.
ACCESS_EXECUTE: bit 6 Execute file (no meaning
for a directory).

The server must return an error if the any access permission cannot
be determined.

IMPLEMENTATION

In general, it is not sufficient for the client to attempt to
deduce access permissions by inspecting the uid, gid, and mode
fields in the file attributes, since the server may perform uid or
gid mapping or enforce additional access control restrictions. It
is also possible that the NFS version 4 protocol server may not be
in the same ID space as the NFS version 4 protocol client. In these
cases (and perhaps others), the NFS version 4 protocol client can
not reliably perform an access check with only current file
attributes.

In the NFS version 2 protocol, the only reliable way to determine
whether an operation was allowed was to try it and see if it
succeeded or failed. Using the ACCESS procedure in the NFS version
4 protocol, the client can ask the server to indicate whether or

Expires: March 2000 [Page 79]

Draft Protocol Specification NFS version 4 September 1999

not one or more classes of operations are permitted. The ACCESS
operation is provided to allow clients to check before doing a

Expires: December 1999 [Page 51]

Draft Protocol Specification NFS version 4 June 1999
series of operations. This is useful in operating systems (such as
UNIX) where permission checking is done only when a directory is
opened. This procedure is also invoked by NFS client access
procedure (called possibly through access(2)). The intent is to
make the behavior of opening a remote file more consistent with the
behavior of opening a local file.

For NFS version 4, the use of the ACCESS procedure when opening a
regular file is deprecated in favor of using OPEN.

The information returned by the server in response to an ACCESS
call is not permanent. It was correct at the exact time that the
server performed the checks, but not necessarily afterwards. The
server can revoke access permission at any time.

The NFS version 4 protocol client should use the effective
credentials of the user to build the authentication information in
the ACCESS request used to determine access rights. It is the
effective user and group credentials that are used in subsequent
read and write operations.

Many implementations do not directly support the ACCESS_DELETE
permission. Operating systems like UNIX will ignore the
ACCESS_DELETE bit if set on an access request on a non-directory
object. In these systems, delete permission on a file is determined
by the access permissions on the directory in which the file
resides, instead of being determined by the permissions of the file
itself. Thus, Therefore the bit mask returned for such a request enumerating which access
rights can be determined will have the ACCESS_DELETE bit value set to 0, indicating that
0. This indicates to the client does not
have this permission. that the server was unable to
check that particular access right. The ACCESS_DELETE bit in the
access mask returned will then be ignored by the client.

ERRORS

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_SERVERFAULT

NFS4ERR_STALE

NFS4ERR_BADHANDLE

Expires: March 2000 [Page 80]

Draft Protocol Specification NFS version 4 September 1999

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 52] 81]

Draft Protocol Specification NFS version 4 June September 1999

10.4. Procedure

12.2.2. Operation 3: CLOSE - Close File

SYNOPSIS

(cfh), stateid -> stateid

ARGUMENT

struct CLOSE4args {
stateid4stateid;
stateid4 stateid;
};

RESULT

union CLOSE4res switch (nfsstat4 status) {
case NFS4_OK:
stateid4stateid;
stateid4 stateid;
default:
void;
};

DESCRIPTION

The CLOSE procedure notifies the server that all share reservations
corresponding to the client supplied stateid should be released.

IMPLEMENTATION

Share reservations for the matching stateid will be released on
successful completion of the CLOSE procedure.

ERRORS

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

Expires: December 1999 March 2000 [Page 53] 82]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_SERVERFAULT

NFS4ERR_EXPIRED

NFS4ERR_GRACE

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 54] 83]

Draft Protocol Specification NFS version 4 June September 1999

10.5. Procedure

12.2.3. Operation 4: COMMIT - Commit Cached Data

SYNOPSIS

(cfh), offset, count -> verifier

ARGUMENT

struct COMMIT4args {
/* CURRENT_FH: file */
offset4offset;
count4count;
offset4 offset;
count4 count;
};

RESULT

struct COMMIT4resok {
writeverf4verf;
writeverf4 verf;
};

union COMMIT4res switch (nfsstat4 status) {
case NFS4_OK:
COMMIT4resokresok4;
COMMIT4resok resok4;
default:
void;
};

DESCRIPTION

The COMMIT procedure forces or flushes data to stable storage that
was previously written with a WRITE operation which had the stable
field set to UNSTABLE4.

The offset provided by the client represents the position within
the file at which the flush is to begin. An offset value of 0
(zero) means to flush data starting at the beginning of the file.
The count as provided by the client is the number of bytes of data
to flush. If count is 0 (zero), a flush from offset to the end of
file is done.

The server returns a write verifier upon successful completion of
the COMMIT. The write verifier is used by the client to determine
if the server has restarted or rebooted between the initial

Expires: December 1999 March 2000 [Page 55] 84]

Draft Protocol Specification NFS version 4 June September 1999

WRITE(s) and the COMMIT. The client does this by comparing the
write verifier returned from the initial writes and the verifier
returned by the COMMIT procedure. The server must vary the value
of the write verifier at each server event that may lead to a loss
of uncommitted data. Most commonly this occurs when the server is
rebooted; however, other events at the server may result in
uncommitted data loss as well.

IMPLEMENTATION

The COMMIT procedure is similar in operation and semantics to the
POSIX fsync(2) system call that synchronizes a file's state with
the disk (file data and metadata is flushed to disk or stable
storage). COMMIT performs the same operation for a client, flushing
any unsynchronized data and metadata on the server to the server's
disk or stable storage for the specified file. Like fsync(2), it
may be that there is some modified data or no modified data to
synchronize. The data may have been synchronized by the server's
normal periodic buffer synchronization activity. COMMIT should
return NFS4_OK, unless there has been an unexpected error.

COMMIT differs from fsync(2) in that it is possible for the client
to flush a range of the file (most likely triggered by a buffer-
reclamation scheme on the client before file has been completely
written).

The server implementation of COMMIT is reasonably simple. If the
server receives a full file COMMIT request, that is starting at
offset 0 and count 0, it should do the equivalent of fsync()'ing
the file. Otherwise, it should arrange to have the cached data in
the range specified by offset and count to be flushed to stable
storage. In both cases, any metadata associated with the file must
be flushed to stable storage before returning. It is not an error
for there to be nothing to flush on the server. This means that
the data and metadata that needed to be flushed have already been
flushed or lost during the last server failure.

The client implementation of COMMIT is a little more complex.
There are two reasons for wanting to commit a client buffer to
stable storage. The first is that the client wants to reuse a
buffer. In this case, the offset and count of the buffer are sent
to the server in the COMMIT request. The server then flushes any
cached data based on the offset and count, and flushes any metadata
associated with the file. It then returns the status of the flush
and the write verifier. The other reason for the client to
generate a COMMIT is for a full file flush, such as may be done at

Expires: December 1999 March 2000 [Page 56] 85]

Draft Protocol Specification NFS version 4 June September 1999

close. In this case, the client would gather all of the buffers
for this file that contain uncommitted data, do the COMMIT
operation with an offset of 0 and count of 0, and then free all of
those buffers. Any other dirty buffers would be sent to the server
in the normal fashion.

After a buffer is written by the client with stable parameter set
to UNSTABLE, the buffer must be considered as modified by the
client until the buffer has either been flushed via a COMMIT
operation or written via a WRITE operation with stable parameter
set to FILE_SYNC or DATA_SYNC. This is done to prevent the buffer
from being freed and reused before the data can be flushed to
stable storage on the server.

When a response comes back from either a WRITE or a COMMIT
operation and it contains a write verifier that is different than
previously returned by the server, the client will need to
retransmit all of the buffers containing uncommitted cached data to
the server. How this is to be done is up to the implementor. If
there is only one buffer of interest, then it should probably be
sent back over in a WRITE request with the appropriate stable
parameter. If there is more than one buffer, it might be worthwhile
retransmitting all of the buffers in WRITE requests with the stable
parameter set to UNSTABLE and then retransmitting the COMMIT
operation to flush all of the data on the server to stable storage.
The timing of these retransmissions is left to the implementor.

The above description applies to page-cache-based systems as well
as buffer-cache-based systems. In those systems, the virtual memory
system will need to be modified instead of the buffer cache.

ERRORS

NFS4ERR_IO

NFS4ERR_LOCKED

NFS4ERR_SERVERFAULT

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 57] 86]

Draft Protocol Specification NFS version 4 June September 1999

10.6. Procedure

12.2.4. Operation 5: CREATE - Create a Non-Regular File Object

SYNOPSIS

(cfh), name, type, how -> (cfh) (cfh), change_info

ARGUMENT

struct CREATE4args {
/* CURRENT_FH: directory for creation */
filename4objname;
fattr4_typetype;
createhow4createhow;
component4 objname;
fattr4_type type;
createhow4 createhow;
};

RESULT

struct change_info4 {
bool atomic;
fattr4_change before;
fattr4_change after;
};

struct CREATE4resok {
change_info4 cinfo;
};

union CREATE4res switch (nfsstat4 status) {
nfsstat4status;
case NFS4_OK:
CREATE4resok resok4;
default:
void;
};

DESCRIPTION

The CREATE procedure creates an non-regular file object in a
directory with a given name. The OPEN procedure MUST be used to
create a regular file.

The need for exclusive create semantics for non-regular

Expires: March 2000 [Page 87]

Draft Protocol Specification NFS version 4 September 1999

files needs to be decided upon and decisions about storage
location of the verifier will need to be determined as
well.

The objtype determines the type of object to be created: directory,
symlink, etc. The how union may have a value of UNCHECKED, GUARDED,
and EXCLUSIVE. UNCHECKED means that the object should be created
without checking for the existence of a duplicate object in the same
directory. In this case, attrbits and attrvals describe the initial
attributes for the file object. GUARDED specifies that the server
should check for the presence of a duplicate object before performing
the create and should fail the request with NFS4ERR_EXIST if a
duplicate object exists. If the object does not exist, the request is

Expires: December 1999 [Page 58]

Draft Protocol Specification NFS version 4 June 1999
performed as described for UNCHECKED. EXCLUSIVE specifies that the
server is to follow exclusive creation semantics, using the verifier
to ensure exclusive creation of the target. No attributes may be
provided in this case, since the server may use the target object
meta-data to store the verifier.

For the directory where the new file object was created, the server
returns change_info4 information in cinfo. With the atomic field of
the change_info4 struct, the server will indicate if the before and
after change attributes were obtained atomically with respect to the
file object creation.

The current filehandle is replaced by that of the new object.

IMPLEMENTATION

The CREATE procedure carries support for EXCLUSIVE create forward
from NFS version 3. As in NFS version 3, this mechanism provides
reliable exclusive creation. Exclusive create is invoked when the
how parameter is EXCLUSIVE. In this case, the client provides a
verifier that can reasonably be expected to be unique. A
combination of a client identifier, perhaps the client network
address, and a unique number generated by the client, perhaps the
RPC transaction identifier, may be appropriate.

If the object does not exist, the server creates the object and
stores the verifier in stable storage. For file systems that do not
provide a mechanism for the storage of arbitrary file attributes,
the server may use one or more elements of the object meta-data to
store the verifier. The verifier must be stored in stable storage
to prevent erroneous failure on retransmission of the request. It
is assumed that an exclusive create is being performed because
exclusive semantics are critical to the application. Because of the

Expires: March 2000 [Page 88]

Draft Protocol Specification NFS version 4 September 1999

expected usage, exclusive CREATE does not rely solely on the
normally volatile duplicate request cache for storage of the
verifier. The duplicate request cache in volatile storage does not
survive a crash and may actually flush on a long network partition,
opening failure windows. In the UNIX local file system
environment, the expected storage location for the verifier on
creation is the meta-data (time stamps) of the object. For this
reason, an exclusive object create may not include initial
attributes because the server would have nowhere to store the
verifier.

If the server can not support these exclusive create semantics,
possibly because of the requirement to commit the verifier to
stable storage, it should fail the CREATE request with the error,
NFS4ERR_NOTSUPP.

During an exclusive CREATE request, if the object already exists,
the server reconstructs the object's verifier and compares it with
the verifier in the request. If they match, the server treats the
request as a success. The request is presumed to be a duplicate of
an earlier, successful request for which the reply was lost and
that the server duplicate request cache mechanism did not detect.

Expires: December 1999 [Page 59]

Draft Protocol Specification NFS version 4 June 1999
If the verifiers do not match, the request is rejected with the
status, NFS4ERR_EXIST.

Once the client has performed a successful exclusive create, it
must issue a SETATTR to set the correct object attributes. Until
it does so, it should not rely upon any of the object attributes,
since the server implementation may need to overload object meta-
data to store the verifier.

Use of the GUARDED attribute does not provide exactly-once
semantics. In particular, if a reply is lost and the server does
not detect the retransmission of the request, the procedure can
fail with NFS4ERR_EXIST, even though the create was performed
successfully.

Note:

1. Need to determine an initial set of attributes
that must be set, and a set of attributes that
can optionally be set, on a per-filetype basis.
For instance, if the filetype is a NF4BLK then
the device attributes must be set.

2. Need to consider the symbolic link path as
an "attribute". No need for a READLINK op

Expires: March 2000 [Page 89]

Draft Protocol Specification NFS version 4 September 1999

if this is so. Similarly, a filehandle could
be defined as an attribute for LINK.

ERRORS

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_EXIST

NFS4ERR_NOTDIR

NFS4ERR_INVAL

NFS4ERR_NOSPC

NFS4ERR_ROFS

NFS4ERR_NAMETOOLONG

NFS4ERR_DQUOT

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 60] 90]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_NAMETOOLONG

NFS4ERR_DQUOT

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

12.2.5. Operation 6: DELEGPURGE - Purge Delegations Awaiting Recovery

SYNOPSIS

clientid ->

ARGUMENT

struct DELEGPURGE4args {
clientid4 clientid;
};

RESULT

struct DELEGPURGE4res {
nfsstat4 status;
};

DESCRIPTION

Purges all of the delegations awaiting recovery for a given client.
This is useful for clients which do not commit delegation
information to stable storage to indicate that conflicting requests
need not be held up awaiting recovery of delegation information.

This operation should also be used by clients which do have
delegation information on stable storage after doing all of
delegation recovery that is needed. Using DELEGPURGE will prevent
any delegations which were made by the server but were not sent to
the client and committed to stable storage from holding up other
clients making conflicting requests.

ERRORS

<TBD>

Expires: December March 2000 [Page 91]

Draft Protocol Specification NFS version 4 September 1999

12.2.6. Operation 7: DELEGRETURN - Return Delegation

SYNOPSIS

stateid ->

ARGUMENT

struct DELEGRETURN4args {
stateid4 stateid;
};

RESULT

struct DELEGRETURN4res {
nfsstat4 status;
};

DESCRIPTION

Returns the delegation represented by the given stateid

ERRORS

<TBD>

Expires: March 2000 [Page 61] 92]

Draft Protocol Specification NFS version 4 June September 1999

10.7. Procedure 6:

12.2.7. Operation 8: GETATTR - Get Attributes

SYNOPSIS

(cfh), attrbits -> attrbits, attrvals

ARGUMENT

struct GETATTR4args {
/* CURRENT_FH: directory or file */
bitmap4 attr_request;
};

RESULT

struct GETATTR4resok {
fattr4 obj_attributes;
};

union GETATTR4res switch (nfsstat4 status) {
case NFS4_OK:
GETATTR4resokresok4;
GETATTR4resok resok4;
default:
void;
};

DESCRIPTION

The GETATTR procedure will obtain attributes from the server. The
client sets a bit in the bitmap argument for each attribute value
that it would like the server to return. The server returns an
attribute bitmap that indicates the attribute values for which it
was able to return, followed by the attribute values ordered lowest
attribute number first.

The server must return a value for each attribute that the client
requests if the attribute is supported by the server. If the
server does not support an attribute or cannot approximate a useful
value then it must not return the attribute value and must not set
the attribute bit in the result bitmap. The server must return an
error if it supports an attribute but cannot obtain its value. In
that case no attribute values will be returned.

Expires: December 1999 March 2000 [Page 62] 93]

Draft Protocol Specification NFS version 4 June September 1999

All servers must support attribute 0 (zero) which is a bitmap of
all supported attributes for the filesystem object.

IMPLEMENTATION

ERRORS

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_JUKEBOX

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 63] 94]

Draft Protocol Specification NFS version 4 June September 1999

10.8. Procedure 7:

12.2.8. Operation 9: GETFH - Get Current Filehandle

SYNOPSIS

(cfh) -> filehandle

ARGUMENT

/* CURRENT_FH: */
void;

RESULT

struct GETFH4resok {
nfs4_fh object;
};

union GETFH4res switch (nfsstat4 status) {
case NFS4_OK:
GETFH4resokresok4;
GETFH4resok resok4;
default:
void;
};

DESCRIPTION

Returns the current filehandle. Operations that change the current
filehandle like LOOKUP or CREATE to not automatically return the
new filehandle as a result. For instance, if a client needs to
lookup a directory entry and obtain its filehandle then the
following request is needed.

1: PUTFH (directory filehandle)
2: LOOKUP (entry name)
3: GETFH

IMPLEMENTATION

Expires: December 1999 March 2000 [Page 64] 95]

Draft Protocol Specification NFS version 4 June September 1999

ERRORS

NFS4ERR_BADHANDLE

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

NFS4ERR_NOFILEHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_STALE

NFS4ERR_WRONGSEC

Expires: December 1999 March 2000 [Page 65] 96]

Draft Protocol Specification NFS version 4 June September 1999

10.9. Procedure 8:

12.2.9. Operation 10: LINK - Create Link to a File

SYNOPSIS

(cfh), directory, newname -> (cfh) (cfh), change_info

ARGUMENT

struct LINK4args {
/* CURRENT_FH: file */
nfs4_fh dir;
filename4
component4 newname;
};

RESULT

struct LINK4resok {
change_info4 cinfo;
};

union LINK4res switch (nfsstat4 status) {
nfsstat4status;
case NFS4_OK:
LINK4resok resok4;
default:
void;
};

DESCRIPTION

The LINK procedure creates an additional newname for the file with
the current filehandle in the new directory dir dir. The current file
handle and link.dir the directory must reside on within the same file system and on
the server.

For the directory, the server returns change_info4 information in
cinfo. With the atomic field of the change_info4 struct, the
server will indicate if the before and after change attributes were
obtained atomically with respect to the link creation.

IMPLEMENTATION

Expires: March 2000 [Page 97]

Draft Protocol Specification NFS version 4 September 1999

Changes to any property of the hard-linked files are reflected in
all of the linked files. When a hard link is made to a file, the
attributes for the file should have a value for nlink that is one
greater than the value before the LINK.

The comments under RENAME regarding object and target residing on
the same file system apply here as well. The comments regarding the
target name applies as well.

ERRORS

NFS4ERR_IO

Expires: December 1999 [Page 66]

Draft Protocol Specification NFS version 4 June 1999

NFS4ERR_ACCES

NFS4ERR_EXIST

NFS4ERR_XDEV

NFS4ERR_NOTDIR

NFS4ERR_INVAL

NFS4ERR_NOSPC

NFS4ERR_ROFS

NFS4ERR_MLINK

NFS4ERR_NAMETOOLONG

NFS4ERR_DQUOT

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 67] 98]

Draft Protocol Specification NFS version 4 June September 1999

10.10. Procedure 9:

12.2.10. Operation 11: LOCK - Create Lock

SYNOPSIS

(cfh) type, seqid, reclaim, owner, offset, length -> stateid,
access

ARGUMENT

struct lockown {
clientid4clientid;
opaqueowner<>;
};

union nfs_lockowner switch (stateid4 stateid) {
case 0:
lockownident;
default:
void;
};

enum nfs4_lock_type {
READ_LT = 1,
WRITE_LT = 2,
READW_LT = 3,/* 3, /* blocking read */
WRITEW_LT = 4/* 4 /* blocking write */
};

struct LOCK4args {
/* CURRENT_FH: file */
nfs4_lock_typetype;
seqid4seqid;
boolreclaim;
nfs_lockownerowner;
offset4offset;
length4length;
nfs4_lock_type type;
seqid4 seqid;
bool reclaim;
stateid4 stateid;
offset4 offset;
length4 length;
};

RESULT

struct lockres {
stateid4stateid;
int32_taccess;
stateid4 stateid;
int32_t access;
};

Expires: December 1999 [Page 68]

Draft Protocol Specification NFS version 4 June 1999

union LOCK4res switch (nfsstat4 status) {
case NFS4_OK:
lockresresult;
lockres result;
default:
void;
};

DESCRIPTION

Expires: March 2000 [Page 99]

Draft Protocol Specification NFS version 4 September 1999

The LOCK procedure requests a record lock for the byte range
specified by the offset and length parameters. The lock type is
also specified to be one of the nfs4_lock_types. If this is a
reclaim request, the reclaim parameter will be TRUE;

IMPLEMENTATION

The File Locking section contains a full description of this and
the other file locking procedures.

ERRORS

NFS4ERR_ACCES

NFS4ERR_ISDIR

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_GRACE

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 69] 100]

Draft Protocol Specification NFS version 4 June September 1999

10.11. Procedure 10:

12.2.11. Operation 12: LOCKT - Test For Lock

SYNOPSIS

(cfh) type, seqid, reclaim, owner, offset, length -> {void,
NFS4ERR_DENIED -> owner}

ARGUMENT

struct LOCK4args {
/* CURRENT_FH: file */
nfs4_lock_typetype;
seqid4seqid;
boolreclaim;
nfs_lockownerowner;
offset4offset;
length4length;
nfs4_lock_type type;
seqid4 seqid;
bool reclaim;
nfs_lockowner owner;
offset4 offset;
length4 length;
};

RESULT

union LOCKT4res switch (nfsstat4 status) {
case NFS4ERR_DENIED:
nfs_lockownerowner;
nfs_lockowner owner;
case NFS4_OK:
void;
default:
void;
};

DESCRIPTION

The LOCKT procedure tests the lock as specified in the argument.
The owner of the lock is returned in the event it is currently
being held; if no lock is held, nothing other than NFS4_OK is
returned.

IMPLEMENTATION

The File Locking section contains a full description of this and
the other file locking procedures.

Expires: December 1999 March 2000 [Page 70] 101]

Draft Protocol Specification NFS version 4 June September 1999

ERRORS

NFS4ERR_ACCES

NFS4ERR_ISDIR

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_DENIED

NFS4ERR_GRACE

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 71] 102]

Draft Protocol Specification NFS version 4 June September 1999

10.12. Procedure 11:

12.2.12. Operation 13: LOCKU - Unlock File

SYNOPSIS

(cfh) type, seqid, reclaim, owner, offset, length -> stateid

ARGUMENT

struct LOCK4args {
/* CURRENT_FH: file */
nfs4_lock_typetype;
seqid4seqid;
boolreclaim;
nfs_lockownerowner;
offset4offset;
length4length;
nfs4_lock_type type;
seqid4 seqid;
bool reclaim;
nfs_lockowner owner;
offset4 offset;
length4 length;
};

RESULT

union LOCKU4res switch (nfsstat4 status) {
caseNFS4_OK:
stateid4stateid_ok;
case NFS4_OK:
stateid4 stateid_ok;
default:
stateid4stateid_oth;
stateid4 stateid_oth;
};

DESCRIPTION

The LOCKU procedure unlocks the record lock specified by the
parameters.

IMPLEMENTATION

The File Locking section contains a full description of this and
the other file locking procedures.

ERRORS

NFS4ERR_ACCES

Expires: December 1999 March 2000 [Page 72] 103]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_ISDIR

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_GRACE

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 73] 104]

Draft Protocol Specification NFS version 4 June September 1999

10.13. Procedure 12:

12.2.13. Operation 14: LOOKUP - Lookup Filename

SYNOPSIS

(cfh), filenames -> (cfh)

ARGUMENT

struct LOOKUP4args {
/* CURRENT_FH: directory */
filename4 filenames<>;
pathname4 path;
};

RESULT

struct LOOKUP4res {
/* CURRENT_FH: object */
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

The current filehandle is assumed to refer to a directory. LOOKUP
evaluates the pathname contained in the array of names and obtains
a new current filehandle from the final name. All but the final
name in the list must be the names of directories.

If the pathname cannot be evaluated either because a component
doesn't exist or because the client doesn't have permission to
evaluate a component of the path, then an error will be returned
and the current filehandle will be unchanged.

IMPLEMENTATION

If the client prefers a partial evaluation of the path then a
sequence of LOOKUP operations can be substituted e.g.

1. PUTFH (directory filehandle)
2. LOOKUP "pub" "foo" "bar"
3. GETFH

Expires: December 1999 March 2000 [Page 74] 105]

Draft Protocol Specification NFS version 4 June September 1999

1. PUTFH (directory filehandle)
2. LOOKUP "pub"
3. GETFH
4. LOOKUP "foo"
5. GETFH
6. LOOKUP "bar"
7. GETFH

NFS version 4 servers depart from the semantics of previous NFS
versions in allowing LOOKUP requests to cross mountpoints on the
server. The client can detect a mountpoint crossing by comparing
the fsid attribute of the directory with the fsid attribute of the
directory looked up. If the fsids are different then the new
directory is a server mountpoint. Unix clients that detect a
mountpoint crossing will need to mount the server's filesystem.

Servers that limit NFS access to "shares" or "exported" filesystems
should provide a pseudo-filesystem into which the exported
filesystems can be integrated, so that clients can browse the
server's namespace. The clients view of a pseudo filesystem will
be limited to paths that lead to exported filesystems.

Note: previous versions of the protocol assigned special semantics
to the names "." and "..". NFS version 4 assigns no special
semantics to these names. The LOOKUPP operator must be used to
lookup a parent directory.

Note that this procedure does not follow symbolic links. The
client is responsible for all parsing of filenames including
filenames that are modified by symbolic links encountered during
the lookup process.

ERRORS

NFS4ERR_NOENT

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_NOTDIR

NFS4ERR_INVAL

NFS4ERR_NAMETOOLONG

Expires: December 1999 March 2000 [Page 75] 106]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_STALE

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 76] 107]

Draft Protocol Specification NFS version 4 June September 1999

10.14. Procedure 13:

12.2.14. Operation 15: LOOKUPP - Lookup Parent Directory

SYNOPSIS

(cfh) -> (cfh)

ARGUMENT

/* CURRENT_FH: object */
void;

RESULT

struct LOOKUPP4res {
/* CURRENT_FH: directory */
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

The current filehandle is assumed to refer to a directory. LOOKUPP
assigns the filehandle for its parent directory to be the current
filehandle. If there is no parent directory an ENOENT error must
be returned. Therefore, ENOENT will be returned by the server when
the current filehandle is at the root or top of the server's file
tree.

IMPLEMENTATION

As for LOOKUP, LOOKUPP will also cross mountpoints.

ERRORS

NFS4ERR_NOENT

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_INVAL

NFS4ERR_STALE

Expires: December 1999 March 2000 [Page 77] 108]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_STALE

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 78] 109]

Draft Protocol Specification NFS version 4 June September 1999

10.15. Procedure 14:

12.2.15. Operation 16: NVERIFY - Verify Difference in Attributes

SYNOPSIS

(cfh), attrbits, attrvals -> -

ARGUMENT

struct NVERIFY4args {
/* CURRENT_FH: object */
bitmap4 attr_request;
fattr4 obj_attributes;
};

RESULT

struct NVERIFY4res {
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

This operation is used to prefix a sequence of operations to be
performed if one or more attributes have changed on some filesystem
object. If all the attributes match then the error NFS4ERR_SAME
must be returned.

IMPLEMENTATION

This operation is useful as a cache validation operator. If the
object to which the attributes belong has changed then the
following operations may obtain new data associated with that
object. For instance, to check if a file has been changed and
obtain new data if it has:

1. PUTFH (public)
2. LOOKUP "pub" "foo" "bar"
3. NVERIFY attrbits attrs
4. READ 0 32767

Expires: December 1999 March 2000 [Page 79] 110]

Draft Protocol Specification NFS version 4 June September 1999

ERRORS

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_SAME

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 80] 111]

Draft Protocol Specification NFS version 4 June September 1999

10.16. Procedure 15:

12.2.16. Operation 17: OPEN - Open a Regular File

SYNOPSIS

(cfh), file, claim, openhow, owner, seqid, reclaim, access, deny -> (cfh),
stateid, access rflags, access, delegation

ARGUMENT

struct OPEN4args {
open_nameorfh file;
open_claim4 claim;
openflag openhow;
nfs_lockowner owner;
seqid4 seqid;
bool reclaim;
int32_t access;
int32_t deny;
};

union open_nameoffh switch (bool reclaim_fh) {
case FALSE:
/* CURRENT_FH: directory */
filename4 filenames<>;
case TRUE:
/* CURRENT_FH: file on reclaim */
void;
};

enum createmode4 {
UNCHECKED = 0,
GUARDED = 1,
EXCLUSIVE = 2
};

union createhow4 switch (createmode4 mode) {
case UNCHECKED:
case GUARDED:
fattr4 createattrs;
case EXCLUSIVE:
createverf4 verf;
};

enum opentype4 {
OPEN4_NOCREATE0,
OPEN4_CREATE1
OPEN4_NOCREATE 0,
OPEN4_CREATE 1
};

union openflag switch (opentype4 opentype) {
case OPEN4_CREATE:
createhow4 how;
default:
void;
};

/*
* Access and Deny constants for open argument
*/

Expires: December March 2000 [Page 112]

Draft Protocol Specification NFS version 4 September 1999

const OPEN4_ACCESS_READ = 0x0001;
const OPEN4_ACCESS_WRITE= 0x0002;
const OPEN4_ACCESS_BOTH = 0x0003;

const OPEN4_DENY_NONE = 0x0000;
const OPEN4_DENY_READ = 0x0001;
const OPEN4_DENY_WRITE = 0x0002;
const OPEN4_DENY_BOTH = 0x0003;

enum open_delegation_type4 {
OPEN_DELEGATE_NONE = 0,
OPEN_DELEGATE_READ = 1,
OPEN_DELEGATE_WRITE = 2
};

enum open_claim_type4 {
CLAIM_NULL = 0,
CLAIM_PREVIOUS = 1,
CLAIM_DELEGATE_CUR = 2,
CLAIM_DELEGATE_PREV = 3
};

struct open_claim_delegate_cur {
pathname4 file;
stateid4 delegate_stateid;
};

union open_claim4 switch (open_claim_type4 claim) {
/*
* No special rights to file. Ordinary OPEN of the specified file.
*/
case CLAIM_NULL:
/* CURRENT_FH: directory */
pathname4 file;

/*
* Right to the file established by an open previous to server
* reboot. File identified by filehandle obtained at that time
* rather than by name.
*/
case CLAIM_PREVIOUS:
/* CURRENT_FH: file being reclaimed */
int32_t delegate_type;

/*
* Right to file based on a delegation granted by the server.
* File is specified by name.
*/

Expires: March 2000 [Page 113]

Draft Protocol Specification NFS version 4 September 1999

case CLAIM_DELEGATE_CUR:
/* CURRENT_FH: directory */
open_claim_delegate_cur delegate_cur_info;

/* Right to file based on a delegation granted to a previous boot
* instance of the client. File is specified by name.
*/
case CLAIM_DELEGATE_PREV:
/* CURRENT_FH: directory */
pathname4 file_delegate_prev;
};

RESULT

/*
* Result flags
*/
/* Mandatory locking is in effect for this file. */
const OPEN4_RESULT_MLOCK = 0x0001;

struct open_read_delegation4 {
stateid4 stateid; /* Stateid for delegation*/
bool recall; /* Pre-recalled flag for
delegations obtained
by reclaim
(CLAIM_PREVIOUS) */
nfsacl4 permissions; /* Defines users who don't
need an ACCESS call to
open for read */
};

struct open_write_delegation4 {
stateid4 stateid; /* Stateid for delegation
be flushed to the server
on close. */
bool recall; /* Pre-recalled flag for
delegations obtained
by reclaim
(CLAIM_PREVIOUS) */
nfs_space_limit4 space_limit; /* Defines condition that
the client must check to
determine whether the
file needs to be flushed
to the server on close.
*/
nfsacl4 permissions; /* Defines users who don't

Expires: March 2000 [Page 81] 114]

Draft Protocol Specification NFS version 4 June September 1999

need an ACCESS call as
part of a delegated
open. */
};

union openflag open_delegation4
switch (opentype4 opentype) (open_delegation_type4 delegation_type) {
case OPEN4_CREATE:
createhow4how;
default: OPEN_DELEGATE_NONE:
void;
case OPEN_DELEGATE_READ:
OPEN4readDelegation read;
case OPEN_DELEGATE_WRITE:
OPEN4writeDelegation write;
};

struct OPEN4resok {
stateid4 stateid; /* Stateid for open */
uint32_t rflags; /* Result flags */
int32_t access; /*
* Access and Deny constants for granted */
open_delegation4 delegation; /* Info on any open argument
delegation */
const OPEN4_ACCESS_READ = 0x0001;
const OPEN4_ACCESS_WRITE= 0x0002;
const OPEN4_ACCESS_BOTH = 0x0003;

const OPEN4_DENY_NONE = 0x0000;
const OPEN4_DENY_READ = 0x0001;
const OPEN4_DENY_WRITE = 0x0002;
const OPEN4_DENY_BOTH = 0x0003;

RESULT
};

union OPEN4res switch (nfsstat4 status) {
case NFS4_OK:
/* CURRENT_FH: opened file */
LOCK4resresok;
OPEN4resok result;
default:
void;
};

DESCRIPTION

The OPEN procedure creates and/or opens a regular file in a
directory with the provided name. If the file does not exist at
the server and creation is desired, specification of the method of
creation is provided by the openhow parameter. The client has the
choice of three creation methods: UNCHECKED, GUARDED, or EXCLUSIVE.

UNCHECKED means that the file should be created without checking
for the existence of a duplicate object in the same directory. For
this type of create, createattrs specifies the initial set of
attributes for the file (NOTE: need to define exactly which
attributes should be set and if the file exists, should the
attributes be modified if the file exists). If GUARDED is
specified, the server checks for the presence of a duplicate object

Expires: December 1999 March 2000 [Page 82] 115]

Draft Protocol Specification NFS version 4 June September 1999

by name before performing the create. If a duplicate exists, an
error of NFS4ERR_EXIST is returned as the status. If the object
does not exist, the request is performed as described for
UNCHECKED.

EXCLUSIVE specifies that the server is to follow exclusive creation
semantics, using the verifier to ensure exclusive creation of the
target. The server should check for the presence of a duplicate
object by name. If the object does not exist, the server creates
the object and stores the verifier with the object. If the object
does exist and the stored verifier matches the client provided
verifier, the server uses the existing object as the newly created
object. If the stored verifier does not match, then an error of
NFS4ERR_EXIST is returned. No attributes may be provided in this
case, since the server may use an attribute of the target object to
store the verifier. (NOTE: does a specific attribute need to be
specified for storage of verifier )

Upon successful creation, the current filehandle is replaced by
that of the new object.

The OPEN procedure provides for DOS SHARE capability with the use
of the access and deny fields of the OPEN arguments. The client
specifies at OPEN the required access and deny modes. For clients
that do not directly support SHAREs (i.e. Unix), the expected deny
value is DENY_NONE. In the case that there is a existing SHARE
reservation that conflicts with the OPEN request, the server
returns the error NFS4ERR_DENIED. For a complete SHARE request,
the client must provide values for the owner and seqid fields for
the OPEN argument. For additional discussion of SHARE semantics
see the section on 'Share Reservations'.

In the case that the client is recovering state from a server
failure, the reclaim field of the OPEN argument is use used to signify
that the request is meant to reclaim state previously held.

The file "claim" field of the OPEN argument allows the client is used to specify an
open by name or by filehandle. The filehandle MAY only the file
to be
specified in opened and the case that state information which the OPEN client claims to
possess. There are four basic claim types which cover the various
situations for an OPEN. They are as follows:

CLAIM_NULL
For the client, this is a reclaim request. It new OPEN
request and there is
expected that if no previous state
associate with the file for the client.

Expires: March 2000 [Page 116]

Draft Protocol Specification NFS version 4 September 1999

CLAIM_PREVIOUS
The client is claiming basic OPEN state
for a file that was held previous to a
server reboot. Generally used when a
server is providing returning persistent filehandles, file
handles; the client will may not have the
file names saved name to request a reclaim the OPEN.

CLAIM_DELEGATE_CUR
The client is claiming a delegation for
OPEN as granted by name. the server.
Generally this is done as part of
recalling a delegation.

CLAIM_DELEGATE_PREV
The client is claiming a delegation
granted to a previous client instance;
used after the client reboots.

For non-reclaim OPEN requests whose claim type is other than CLAIM_PREVIOUS
(i.e. requests other than those devoted to reclaiming opens after a
server reboot) that reach the server during its grace or lease
expiration period, the server returns an error of NFS4ERR_GRACE.

Expires: December 1999 [Page 83]

Draft Protocol Specification NFS version 4 June 1999

For any OPEN request, the server may return an open delegation,
which allows further opens and closes to be handled locally on the
client as described in the section Open Delegation. Note that
delegation is up to the server to decide. The client should never
assume that delegation will or will not be granted in a particular
instance. It should always be prepared for either case. A partial
exception is the reclaim (CLAIM_PREVIOUS) case, in which a
delegation type is claimed. In this case, delegation will always
be granted, although the server may specify an immediate recall in
the delegation structure.

IMPLEMENTATION

The OPEN procedure contains support for EXCLUSIVE create. The
mechanism is similar to the support in NFS version 3 [RFC1813]. As
in NFS version 3, this mechanism provides reliable exclusive
creation. Exclusive create is invoked when the how parameter is
EXCLUSIVE. In this case, the client provides a verifier that can
reasonably be expected to be unique. A combination of a client
identifier, perhaps the client network address, and a unique number
generated by the client, perhaps the RPC transaction identifier,
may be appropriate.

If the object does not exist, the server creates the object and

Expires: March 2000 [Page 117]

Draft Protocol Specification NFS version 4 September 1999

stores the verifier in stable storage. For file systems that do not
provide a mechanism for the storage of arbitrary file attributes,
the server may use one or more elements of the object meta-data to
store the verifier. The verifier must be stored in stable storage
to prevent erroneous failure on retransmission of the request. It
is assumed that an exclusive create is being performed because
exclusive semantics are critical to the application. Because of the
expected usage, exclusive CREATE does not rely solely on the
normally volatile duplicate request cache for storage of the
verifier. The duplicate request cache in volatile storage does not
survive a crash and may actually flush on a long network partition,
opening failure windows. In the UNIX local file system
environment, the expected storage location for the verifier on
creation is the meta-data (time stamps) of the object. For this
reason, an exclusive object create may not include initial
attributes because the server would have nowhere to store the
verifier.

If the server can not support these exclusive create semantics,
possibly because of the requirement to commit the verifier to
stable storage, it should fail the OPEN request with the error,
NFS4ERR_NOTSUPP.

Once the client has performed a successful exclusive create, it
must issue a SETATTR to set the correct object attributes. Until
it does so, it should not rely upon any of the object attributes,

Expires: December 1999 [Page 84]

Draft Protocol Specification NFS version 4 June 1999
since the server implementation may need to overload object meta-
data to store the verifier. The subsequent SETATTR must not occur
in the same COMPOUND request as the OPEN. This separation will
guarantee that the exclusive create mechanism will continue to
function properly in the face of retransmission of the request.

For SHARE reservations, the client must specify a value for access

Expires: March 2000 [Page 118]

Draft Protocol Specification NFS version 4 September 1999

that is one of READ, WRITE, or BOTH. For deny, the client must
specify one of NONE, READ, WRITE, or BOTH. If the client fails to
do this, the server must return NFS4ERR_INVAL.

The OPEN call

ERRORS

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_EXIST

NFS4ERR_NOTDIR

NFS4ERR_NOSPC

NFS4ERR_ROFS

NFS4ERR_NAMETOOLONG

NFS4ERR_DQUOT

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_SHARE_DENIED

NFS4ERR_GRACE

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 85] 119]

Draft Protocol Specification NFS version 4 June September 1999

10.17. Procedure 16:

12.2.17. Operation 18: OPENATTR - Open Named Attribute Directory

SYNOPSIS

(cfh) -> (cfh)

ARGUMENT

/* CURRENT_FH: file or directory */
void;

RESULT

struct OPENATTR4res {
/* CURRENT_FH: name attr directory*/
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

The OPENATTR procedure is used to obtain the filehandle of the
named attribute directory associated with the current filehandle.
The result of the OPENATTR will be a filehandle of type NF4ATTRDIR.
From this filehandle, READDIR and LOOKUP procedures can be used to
obtain filehandles for the various named attributes associated with
the original file system object. Filehandles returned within the
named attribute directory will have a type of NF4NAMEDATTR.

IMPLEMENTATION

If the server does not support named attributes for the current
filehandle, an error of NFS4ERR_NOTSUPP will be returned to the
client.

ERRORS

NFS4ERR_NOENT

NFS4ERR_IO

Expires: December 1999 March 2000 [Page 86] 120]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_ACCES

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_JUKEBOX

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 87] 121]

Draft Protocol Specification NFS version 4 June September 1999

10.18. Procedure 17:

12.2.18. Operation 19: PUTFH - Set Current Filehandle

SYNOPSIS

filehandle -> (cfh)

ARGUMENT

struct PUTFH4args {
nfs4_fh object;
};

RESULT

struct PUTFH4res {
/* CURRENT_FH: */
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

Replaces the current filehandle with the filehandle provided as an
argument.

IMPLEMENTATION

Commonly used as the first operator in any NFS request to set the
context for following operations.

ERRORS

NFS4ERR_BADHANDLE

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

NFS4ERR_SERVERFAULT

NFS4ERR_WRONGSEC

Expires: December March 2000 [Page 122]

Draft Protocol Specification NFS version 4 September 1999

NFS4ERR_STALE

NFS4ERR_WRONGSEC

Expires: March 2000 [Page 88] 123]

Draft Protocol Specification NFS version 4 June September 1999

10.19. Procedure 18:

12.2.19. Operation 20: PUTPUBFH - Set Public Filehandle

SYNOPSIS

- -> (cfh)

ARGUMENT

void;

RESULT

struct PUTPUBFH4res {
/* CURRENT_FH: root fh */
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

Replaces the current filehandle with the filehandle that represents
the public filehandle of the server's namespace. This filehandle
may be different from the "root" filehandle which may be associated
with some other directory on the server.

IMPLEMENTATION

Used as the first operator in any NFS request to set the context
for following operations.

ERRORS

NFS4ERR_SERVERFAULT NFS4ERR_WRONGSEC

Expires: December 1999 March 2000 [Page 89] 124]

Draft Protocol Specification NFS version 4 June September 1999

10.20. Procedure 19:

12.2.20. Operation 21: PUTROOTFH - Set Root Filehandle

SYNOPSIS

- -> (cfh)

ARGUMENT

void;

RESULT

struct PUTROOTFH4res {
/* CURRENT_FH: root fh */
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

Replaces the current filehandle with the filehandle that represents
the root of the server's namespace. From this filehandle a LOOKUP
operation can locate any other filehandle on the server. This
filehandle may be different from the "public" filehandle which may
be associated with some other directory on the server.

IMPLEMENTATION

Commonly used as the first operator in any NFS request to set the
context for following operations.

ERRORS

NFS4ERR_SERVERFAULT

NFS4ERR_WRONTSEC

Expires: December 1999 March 2000 [Page 90] 125]

Draft Protocol Specification NFS version 4 June September 1999

10.21. Procedure 20:

12.2.21. Operation 22: READ - Read from File

SYNOPSIS

(cfh), offset, count, stateid -> eof, data

ARGUMENT

struct READ4args {
/* CURRENT_FH: file */
stateid4stateid;
offset4offset;
count4count;
stateid4 stateid;
offset4 offset;
count4 count;
};

RESULT

struct READ4resok {
booleof;
opaquedata<>;
bool eof;
opaque data<>;
};

union READ4res switch (nfsstat4 status) {
case NFS4_OK:
READ4resokresok4;
READ4resok resok4;
default:
void;
};

DESCRIPTION

The READ procedure reads data from the regular file identified by
the current filehandle.

The client provides an offset of where the READ is to start and a
count of how many bytes are to be read. An offset of 0 (zero)
means to read data starting at the beginning of the file. If offset
is greater than or equal to the size of the file, the status,
NFS4_OK, is returned with a data length set to 0 (zero) and eof set
to TRUE. The READ is subject to access permissions checking.

If the client specifies a count value of 0 (zero), the READ

Expires: December 1999 March 2000 [Page 91] 126]

Draft Protocol Specification NFS version 4 June September 1999

succeeds and returns 0 (zero) bytes of data again subject to access
permissions checking. The server may choose to return fewer bytes
than specified by the client. The client needs to check for this
condition and handle the condition appropriately.

The stateid value for a READ request represents a value returned
from a previous record lock or share reservation request. Used by
the server to verify that the associated lock is still valid and to
update lease timeouts for the client.

If the read ended at the end-of-file (formally, in a correctly
formed READ request, if offset + count is equal to the size of the
file), eof is returned as TRUE; otherwise it is FALSE. A successful
READ of an empty file will always return eof as TRUE.

IMPLEMENTATION

It is possible for the server to return fewer than count bytes of
data. If the server returns less than the count requested and eof
set to FALSE, the client should issue another READ to get the
remaining data. A server may return less data than requested under
several circumstances. The file may have been truncated by another
client or perhaps on the server itself, changing the file size from
what the requesting client believes to be the case. This would
reduce the actual amount of data available to the client. It is
possible that the server may back off the transfer size and reduce
the read request return. Server resource exhaustion may also occur
necessitating a smaller read return.

If the file is locked the server will return an NFS4ERR_LOCKED
error. Since the lock may be of short duration, the client may
choose to retransmit the READ request (with exponential backoff)
until the operation succeeds.

ERRORS

NFS4ERR_IO

NFS4ERR_NXIO

NFS4ERR_ACCES

NFS4ERR_INVAL

NFS4ERR_STALE

Expires: December 1999 March 2000 [Page 92] 127]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_BADHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_DENIED

NFS4ERR_JUKEBOX

NFS4ERR_EXPIRED

NFS4ERR_LOCKED

NFS4ERR_GRACE

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 93] 128]

Draft Protocol Specification NFS version 4 June September 1999

10.22. Procedure 21:

12.2.22. Operation 23: READDIR - Read Directory

SYNOPSIS
(cfh), cookie, dircount, maxcount, attrbits -> { cookie, filename,
attrbits, attributes }

ARGUMENT

struct READDIR4args {
/* CURRENT_FH: directory */
nfs_cookie4cookie;
count4dircount;
count4maxcount;
bitmap4attr_request;
nfs_cookie4 cookie;
count4 dircount;
count4 maxcount;
bitmap4 attr_request;

};

RESULT

struct entry4 {
nfs_cookie4cookie;
filename4name;
fattr4attrs;
entry4*nextentry;
nfs_cookie4 cookie;
component4 name;
fattr4 attrs;
entry4 *nextentry;
};

struct dirlist4 {
entry4*entries;
booleof;
entry4 *entries;
bool eof;
};

struct READDIR4resok {
dirlist4reply;
dirlist4 reply;
};

union READDIR4res switch (nfsstat4 status) {
case NFS4_OK:
READDIR4resokresok4;
READDIR4resok resok4;
default:
void;
};

Expires: December 1999 March 2000 [Page 94] 129]

Draft Protocol Specification NFS version 4 June September 1999

DESCRIPTION

The READDIR procedure retrieves a variable number of entries from a
file system directory and returns complete information about each
entry along with information to allow the client to request
additional directory entries in a subsequent READDIR.

The arguments contain a cookie value that represents where the
READDIR should start within the directory. A value of 0 (zero) for
the cookie is used to start reading at the beginning of the
directory. For subsequent READDIR requests, the client specifies a
cookie value that is provided by the server on a previous READDIR
request.

The dircount portion of the argument is the maximum number of bytes
of directory information that should be returned. This value does
not include the size of attributes or filehandle values that may be
returned in the result.

The maxcount value of the argument specifies the maximum number of
bytes for the result. This maximum size represents all of the data
being returned and includes the XDR overhead. The server may
return less data.

Finally, attrbits represents the list of attributes the client
wants returned for each directory entry supplied by the server.

On successful return, the server's response will provide a list of
directory entries. Each of these entries contains the name of the
directory entry, a cookie value for that entry and the associated
attributes as requested. The cookie value is only meaningful to
the server and is used as a "bookmark" for the directory entry. As
mentioned, this cookie is used by the client for subsequent READDIR
operations so that it may continue reading a directory. The cookie
is similar in concept to a READ offset but should not be
interpreted as such by the client. Ideally, the cookie value
should not change if the directory is modified.

IMPLEMENTATION

Issues that need to be understood for this procedure include
increased cache flushing activity on the client (as new file
handles are returned with names which are entered into caches) and
over-the-wire overhead versus expected subsequent LOOKUP and
GETATTR elimination.

The dircount and maxcount fields are included as an optimization.

Expires: December 1999 March 2000 [Page 95] 130]

Draft Protocol Specification NFS version 4 June September 1999

Consider a READDIR call on a UNIX operating system implementation
for 1048 bytes; the reply does not contain many entries because of
the overhead due to attributes and file handles. An alternative is
to issue a READDIR call for 8192 bytes and then only use the first
1048 bytes of directory information. However, the server doesn't
know that all that is needed is 1048 bytes of directory information
(as would be returned by READDIR). It sees the 8192 byte request
and issues a VOP_READDIR for 8192 bytes. It then steps through all
of those directory entries, obtaining attributes and file handles
for each entry. When it encodes the result, the server only
encodes until it gets 8192 bytes of results which include the
attributes and file handles. Thus, it has done a larger VOP_READDIR
and many more attribute fetches than it needed to. The ratio of the
directory entry size to the size of the attributes plus the size of
the file handle is usually at least 8 to 1. The server has done
much more work than it needed to.

The solution to this problem is for the client to provide two
counts to the server. The first is the number of bytes of directory
information that the client really wants, dircount. The second is
the maximum number of bytes in the result, including the attributes
and file handles, maxcount. Thus, the server will issue a
VOP_READDIR for only the number of bytes that the client really
wants to get, not an inflated number. This should help to reduce
the size of VOP_READDIR requests on the server, thus reducing the
amount of work done there, and to reduce the number of VOP_LOOKUP,
VOP_GETATTR, and other calls done by the server to construct
attributes and file handles.

ERRORS

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_NOTDIR

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_BAD_COOKIE

NFS4ERR_TOOSMALL

Expires: December 1999 March 2000 [Page 96] 131]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_JUKEBOX

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 97] 132]

Draft Protocol Specification NFS version 4 June September 1999

10.23. Procedure 22:

12.2.23. Operation 24: READLINK - Read Symbolic Link

SYNOPSIS

(cfh) -> linktext

ARGUMENT

/* CURRENT_FH: symlink */
void;

RESULT

struct READLINK4resok {
linktext4link;
linktext4 link;
};

union READLINK4res switch (nfsstat4 status) {
case NFS4_OK:
READLINK4resokresok4;
READLINK4resok resok4;
default:
void;
};

DESCRIPTION

READLINK reads the data associated with a symbolic link. The data
is a UTF-8 string that is opaque to the server. That is, whether
created by an NFS client or created locally on the server, the data
in a symbolic link is not interpreted when created, but is simply
stored.

IMPLEMENTATION

A symbolic link is nominally a pointer to another file. The data
is not necessarily interpreted by the server, just stored in the
file. It is possible for a client implementation to store a path
name that is not meaningful to the server operating system in a
symbolic link. A READLINK operation returns the data to the client
for interpretation. If different implementations want to share
access to symbolic links, then they must agree on the

Expires: December 1999 March 2000 [Page 98] 133]

Draft Protocol Specification NFS version 4 June September 1999

interpretation of the data in the symbolic link.

The READLINK operation is only allowed on objects of type, NF4LNK.
The server should return the error, NFS4ERR_INVAL, if the object is
not of type, NF4LNK.

ERRORS

NFS4ERR_IO

NFS4ERR_INVAL

NFS4ERR_ACCES

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_JUKEBOX

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 99] 134]

Draft Protocol Specification NFS version 4 June September 1999

10.24. Procedure 23:

12.2.24. Operation 25: REMOVE - Remove Filesystem Object

SYNOPSIS

(cfh), filename -> - change_info

ARGUMENT

struct REMOVE4args {
/* CURRENT_FH: directory */
filename4target;
component4 target;
};

RESULT

struct REMOVE4resok {
change_info4 cinfo;
}

union REMOVE4res switch (nfsstat4 status) {
nfsstat4status;
};
case NFS4_OK:
REMOVE4resok resok4;
default:
void;
}

DESCRIPTION

The REMOVE procecure removes (deletes) a directory entry named by
filename from the directory corresponding to the current
filehandle. If the entry in the directory was the last reference
to the corresponding file system object, the object may be
destroyed.

For the directory where the filename was removed, the server
returns change_info4 information in cinfo. With the atomic field
of the change_info4 struct, the server will indicate if the before
and after change attributes were obtained atomically with respect
to the removal.

Expires: March 2000 [Page 135]

Draft Protocol Specification NFS version 4 September 1999

IMPLEMENTATION

NFS versions 2 and 3 required a different operator RMDIR for
directory removal. NFS version 4 REMOVE can be used to delete any
directory entry independent of its filetype.

The concept of last reference is server specific. However, if the
nlink field in the previous attributes of the object had the value
1, the client should not rely on referring to the object via a file
handle. Likewise, the client should not rely on the resources (disk
space, directory entry, and so on.) formerly associated with the
object becoming immediately available. Thus, if a client needs to
be able to continue to access a file after using REMOVE to remove
it, the client should take steps to make sure that the file will

Expires: December 1999 [Page 100]

Draft Protocol Specification NFS version 4 June 1999
still be accessible. The usual mechanism used is to use RENAME to
rename the file from its old name to a new hidden name.

ERRORS

NFS4ERR_NOENT

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_NOTDIR

NFS4ERR_ROFS

NFS4ERR_NAMETOOLONG

NFS4ERR_NOTEMPTY

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 101] 136]

Draft Protocol Specification NFS version 4 June September 1999

10.25. Procedure 24:

12.2.25. Operation 26: RENAME - Rename Directory Entry

SYNOPSIS

(cfh), oldname, newdir, newname -> - source_change_info,
target_change_info

ARGUMENT

struct RENAME4args {
/* CURRENT_FH: source directory */
filename4oldname;
nfs4_fhnewdir;
filename4newname;
component4 oldname;
nfs4_fh newdir;
component4 newname;
};

RESULT

struct RENAME4resok {
change_info4 source_cinfo;
change_info4 target_cinfo;
};

union RENAME4res switch (nfsstat4 status) {
nfsstat4status;
case NFS4_OK:
RENAME4resok resok4;
default:
void;
};

DESCRIPTION

RENAME renames the object identified by oldname in the directory
corresponding to the current filehandle to newname in directory
newdir. The operation is required to be atomic to the client.
Source and target directories must reside on the same file system
on the server.

If the directory, newdir, already contains an entry with the name,
newname, the source object must be compatible with the target:
either both are non-directories or both are directories and the
target must be empty. If compatible, the existing target is removed

Expires: March 2000 [Page 137]

Draft Protocol Specification NFS version 4 September 1999

before the rename occurs. If they are not compatible or if the
target is a directory but not empty, the server should return the
error, NFS4ERR_EXIST.

If oldname and newname both refer to the same file (they might be
hard links of each other), then RENAME should perform no action and
return success.

For both directories involved in the RENAME, the server returns
change_info4 information. With the atomic field of the
change_info4 struct, the server will indicate if the before and
after change attributes were obtained atomically with respect to
the rename.

IMPLEMENTATION

The RENAME operation must be atomic to the client. The statement
"source and target directories must reside on the same file system

Expires: December 1999 [Page 102]

Draft Protocol Specification NFS version 4 June 1999
on the server" means that the fsid fields in the attributes for the
directories are the same. If they reside on different file systems,
the error, NFS4ERR_XDEV, is returned. Even though the operation is
atomic, the status, NFS4ERR_MLINK, may be returned if the server
used a "unlink/link/unlink" sequence internally.

A file handle may or may not become stale on a rename. However,
server implementors are strongly encouraged to attempt to keep file
handles from becoming stale in this fashion.

On some servers, the filenames, "." and "..", are illegal as either
oldname or newname. In addition, neither oldname nor newname can be
an alias for the source directory. These servers will return the
error, NFS4ERR_INVAL, in these cases.

If oldname and newname both refer to the same file (they might be
hard links of each other), then RENAME should perform no action and
return success.

ERRORS

NFS4ERR_NOENT

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_EXIST

NFS4ERR_XDEV

NFS4ERR_NOTDIR

NFS4ERR_ISDIR

NFS4ERR_INVAL

NFS4ERR_NOSPC

NFS4ERR_ROFS

NFS4ERR_MLINK

NFS4ERR_NAMETOOLONG

NFS4ERR_NOTEMPTY

Expires: December 1999 March 2000 [Page 103] 138]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_NOTDIR

NFS4ERR_ISDIR

NFS4ERR_INVAL

NFS4ERR_NOSPC

NFS4ERR_ROFS

NFS4ERR_MLINK

NFS4ERR_NAMETOOLONG

NFS4ERR_NOTEMPTY

NFS4ERR_DQUOT

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 104] 139]

Draft Protocol Specification NFS version 4 June September 1999

10.26. Procedure 25:

12.2.26. Operation 27: RENEW - Renew a Lease

SYNOPSIS

stateid -> ()

ARGUMENT

struct RENEW4args {
stateid4stateid;
stateid4 stateid;
};

RESULT

struct RENEW4res {
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

The RENEW procedure is used by the client to renew leases which it
currently holds at a server. The processing the RENEW request, the
server renews all leases associated with the client. The
associated leases are determined by the client id provided via the
SETCLIENTID procedure.

IMPLEMENTATION

ERRORS

NFS4ERR_SERVERFAULT

NFS4ERR_EXPIRED

NFS4ERR_GRACE

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 105] 140]

Draft Protocol Specification NFS version 4 June September 1999

10.27. Procedure 25:

12.2.27. Operation 28: RESTOREFH - Restore Saved Filehandle

SYNOPSIS

(sfh) -> (cfh)

ARGUMENT

/* SAVED_FH: */
void;

RESULT

struct RESTOREFH4res {
/* CURRENT_FH: value of saved fh */
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

Set the current filehandle to the value in the saved filehandle.
If there is no saved filehandle then return an error NFS4ERR_INVAL.

IMPLEMENTATION

Procedures like OPEN and LOOKUP use the current filehandle to
represent a directory and replace it with a new filehandle.
Assuming the previous filehandle was saved with a SAVEFH operator,
the previous filehandle can be restored as the current filehandle.
This is commonly used to obtain post-operation attributes for the
directory, e.g.

1. PUTFH (directory filehandle)
2. SAVEFH
3. GETATTR attrbits (pre-op dir attrs)
4. CREATE optbits "foo" attrs
5. GETATTR attrbits (file attributes)
6. RESTOREFH
7. GETATTR attrbits (post-op dir attrs)

Expires: December 1999 March 2000 [Page 106] 141]

Draft Protocol Specification NFS version 4 June September 1999

ERRORS

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

NFS4ERR_NOFILEHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_STALE

NFS4ERR_WRONGSEC

Expires: December 1999 March 2000 [Page 107] 142]

Draft Protocol Specification NFS version 4 June September 1999

10.28. Procedure 27:

12.2.28. Operation 29: SAVEFH - Save Current Filehandle

SYNOPSIS

(cfh) -> (sfh)

ARGUMENT

/* CURRENT_FH: */
void;

RESULT

struct SAVEFH4res {
/* SAVED_FH: value of current fh */
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

Save the current filehandle. If a previous filehandle was saved
then it is no longer accessible. The saved filehandle can be
restored as the current filehandle with the RESTOREFH operator.

IMPLEMENTATION

ERRORS

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

NFS4ERR_NOFILEHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_STALE

Expires: December March 2000 [Page 143]

Draft Protocol Specification NFS version 4 September 1999

NFS4ERR_WRONGSEC

Expires: March 2000 [Page 108] 144]

Draft Protocol Specification NFS version 4 June September 1999

10.29. Procedure 28:

12.2.29. Operation 30: SECINFO - Obtain Available Security

SYNOPSIS

(cfh), filename -> { secinfo }

ARGUMENT

struct SECINFO4args {
/* CURRENT_FH: */
filename4name;
component4 name;
};

RESULT

struct rpc_flavor_info rpcsec_gss_info {
sec_oid4oid;
qop4qop;
sec_oid4 oid;
qop4 qop;
rpc_gss_svc_t service;
};

struct secinfo4 {
rpc_flavor4flavor;
rpc_flavor_info *flavor_info;
secinfo4*nextentry;
unsigned int flavor;
opaque flavor_info<>; /* null for AUTH_SYS, AUTH_NONE;
contains rpcsec_gss_info for
RPCSEC_GSS. */
};

struct SECINFO4resok {
secinfo4reply;
secinfo4 reply<>;
};

union SECINFO4res switch (nfsstat4 status) {
case NFS4_OK:
SECINFO4resokresok4;
SECINFO4resok resok4;
default:
void;
};

DESCRIPTION

Expires: March 2000 [Page 145]

Draft Protocol Specification NFS version 4 September 1999

The SECINFO procedure is used by the client to obtain a list of

Expires: December 1999 [Page 109]

Draft Protocol Specification NFS version 4 June 1999
valid RPC authentication flavors for a specific file handle, file
name pair. The result will contain an array which represents the
security mechanisms available. The array entries are represented
by the secinfo4 structure. The field 'flavor' will contain a value
of AUTH_NONE, AUTH_SYS (as defined in [RFC1831]), or RPCSEC_GSS (as
defined in [RFC2203]).

For the flavors, AUTH_NONE, AUTH_SYS, AUTH_DH, and
AUTH_KRB4 AUTH_SYS no additional security
information is returned. For a return value of AUTH_RPCSEC_GSS, RPCSEC_GSS, a
security triple is returned that contains the mechanism object id
(as defined in [RFC2078]), the quality of protection (as defined in [RFC 2078])
[RFC2078]) and the service type (as defined in [RFC2203]). It is
possible for SECINFO to return multiple entries with flavor equal
to AUTH_RPCSEC_GSS RPCSEC_GSS with different security triple values.

IMPLEMENTATION

The SECINFO procedure is expected to be used by the NFS client when
the error value of NFS4ERR_WRONGSEC is returned from another NFS
procedure. This signifies to the client that the server's security
policy is different from what the client is currently using. At
this point, the client is expected to obtain a list of possible
security flavors and choose what best suits its policies.

ERRORS

NFS4ERR_SERVERFAULT

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 110] 146]

Draft Protocol Specification NFS version 4 June September 1999

10.30. Procedure 29:

12.2.30. Operation 31: SETATTR - Set Attributes

SYNOPSIS

(cfh), attrbits, attrvals -> -

ARGUMENT

struct SETATTR4args {
/* CURRENT_FH: target object */
stateid4 stateid;
fattr4 obj_attributes;
};

RESULT

struct SETATTR4res {
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

The SETATTR Procedure changes one or more of the attributes of a
file system object. The new attributes are specified with a bitmap
and the attributes that follow the bitmap in bit order.

The stateid is necessary for SETATTR's that change the size of file
(modify the attribute object_size). This stateid represents a
record lock, share reservation, or delegation which must be valid
for the SETATTR to modify the file data.

IMPLEMENTATION

The file size attribute is used to request changes to the size of a
file. A value of 0 (zero) causes the file to be truncated, a value
less than the current size of the file causes data from new size to
the end of the file to be discarded, and a size greater than the
current size of the file causes logically zeroed data bytes to be
added to the end of the file. Servers are free to implement this
using holes or actual zero data bytes. Clients should not make any

Expires: March 2000 [Page 147]

Draft Protocol Specification NFS version 4 September 1999

assumptions regarding a server's implementation of this feature,
beyond that the bytes returned will be zeroed. Servers must support
extending the file size via SETATTR.

SETATTR is not guaranteed atomic. A failed SETATTR may partially
change a file's attributes.

Expires: December 1999 [Page 111]

Draft Protocol Specification NFS version 4 June 1999

Changing the size of a file with SETATTR indirectly changes the
time_modify. A client must account for this as size changes can
result in data deletion.

If server and client times differ, programs that compare client
time to file times can break. A time maintenance protocol should be
used to limit client/server time skew.

If the server cannot successfully set all the attributes it must
return an NFS4ERR_INVAL error. If the server can only support 32
bit offsets and sizes, a SETATTR request to set the size of a file
to larger than can be represented in 32 bits will be rejected with
this same error.

ERRORS

NFS4ERR_PERM

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_INVAL

NFS4ERR_FBIG

NFS4ERR_NOSPC

NFS4ERR_ROFS

NFS4ERR_DQUOT

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_JUKEBOX

NFS4ERR_DENIED

NFS4ERR_GRACE

NFS4ERR_FHEXPIRED

Expires: December 1999 March 2000 [Page 112] 148]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_JUKEBOX

NFS4ERR_DENIED

NFS4ERR_GRACE

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 113] 149]

Draft Protocol Specification NFS version 4 June September 1999

10.31. Procedure 30:

12.2.31. Operation 32: SETCLIENTID - Negotiated Clientid

SYNOPSIS

verifier, client -> clientid

ARGUMENT

struct cid {
opaque verifier<4>; verifier[4];
opaque id<>;
};

union nfs_client_id switch (clientid4 clientid) {
case 0:
cidident;
cid ident;
default:
void;
};

struct SETCLIENTID4args {
seqid4seqid;
nfs_client_idclient;
seqid4 seqid;
nfs_client_id client;
};

RESULT

union SETCLIENTID4res switch (nfsstat4 status) {
case NFS4_OK:
clientid4clientid;
clientid4 clientid;
default:
void;
};

DESCRIPTION

The SETCLIENTID procedure introduces the ability of the client to
notify the server of its intention to use a particular client
identifier and verifier pair. Upon successful completion the
server will return a clientid which is used in subsequent file
locking requests.

Expires: December 1999 March 2000 [Page 114] 150]

Draft Protocol Specification NFS version 4 June September 1999

IMPLEMENTATION

The server takes the verifier and client identification supplied
and search for a match of the client identification. If no match
is found the server saves the principal/uid information along with
the verifier and client identification and returns a unique
clientid that is used as a short hand reference to the supplied
information.

If the server find matching client identification and a
corresponding match in principal/uid, the server releases all
locking state for the client and returns a new clientid.

ERRORS

NFS4ERR_INVAL

NFS4ERR_SERVERFAULT

NFS4ERR_CLID_INUSE

Expires: December 1999 March 2000 [Page 115] 151]

Draft Protocol Specification NFS version 4 June September 1999

10.32. Procedure 31:

12.2.32. Operation 33: VERIFY - Verify Same Attributes

SYNOPSIS

(cfh), attrbits, attrvals -> -

ARGUMENT

struct VERIFY4args {
/* CURRENT_FH: object */
bitmap4 attr_request;
fattr4 obj_attributes;
};

RESULT

struct VERIFY4res {
nfsstat4status;
nfsstat4 status;
};

DESCRIPTION

The VERIFY procedure is used to verify that attributes have a value
assumed by the client before proceeding with following operations
in the compound request. For instance, a VERIFY can be used to
make sure that the file size has not changed for an append-mode
write:

1. PUTFH 0x0123456
2. VERIFY attrbits attrs
3. WRITE 450328 4096

If the attributes are not as expected, then the request fails and
the data is not appended to the file.

IMPLEMENTATION

ERRORS

Expires: December 1999 March 2000 [Page 116] 152]

Draft Protocol Specification NFS version 4 June September 1999

NFS4ERR_ACCES

NFS4ERR_INVAL

NFS4ERR_STALE

NFS4ERR_BADHANDLE

NFS4ERR_NOTSUPP

NFS4ERR_SERVERFAULT

NFS4ERR_JUKEBOX

NFS4ERR_FHEXPIRED

NFS4ERR_MOVED

Expires: December 1999 March 2000 [Page 117] 153]

Draft Protocol Specification NFS version 4 June September 1999

10.33. Procedure 32:

12.2.33. Operation 34: WRITE - Write to File

SYNOPSIS

(cfh), offset, count, stability, stateid, data -> count, committed,
verifier

ARGUMENT

enum stable_how4 {
UNSTABLE4 = 0,
DATA_SYNC4 = 1,
FILE_SYNC4 = 2
};

struct WRITE4args {
/* CURRENT_FH: file */
stateid4stateid;
offset4offset;
count4count;
stable_how4stable;
opaquedata<>;
stateid4 stateid;
offset4 offset;
count4 count;
stable_how4 stable;
opaque data<>;
};

RESULT

struct WRITE4resok {
count4count;
stable_how4committed;
writeverf4verf;
count4 count;
stable_how4 committed;
writeverf4 verf;
};

union WRITE4res switch (nfsstat4 status) {
case NFS4_OK:
WRITE4resokresok4;
WRITE4resok resok4;
default:
void;
};

DESCRIPTION

The WRITE procedure is used to write data to a regular file. The regular file. The

Expires: March 2000 [Page 154]

Draft Protocol Specification NFS version 4 September 1999

target file is specified by the current filehandle. The offset
specifies the offset where the data should be written. An offset
of 0 (zero) specifies that the write should start at the beginning
of the file. The count represents the number of bytes of data that
are to be written. If the count is 0 (zero), the WRITE will
succeed and return a count of 0 (zero) subject to permissions
checking. The server may choose to write fewer bytes than
requested by the client.

Part of the write request is a specification of how the write is to
be performed. The client specifies with the stable parameter the
method of how the data is to be processed by the server. If stable
is FILE_SYNC, the server must commit the data written plus all file
system metadata to stable storage before returning results. This
corresponds to the NFS version 2 protocol semantics. Any other
behavior constitutes a protocol violation. If stable is DATA_SYNC,
then the server must commit all of the data to stable storage and
enough of the metadata to retrieve the data before returning. The
server implementor is free to implement DATA_SYNC in the same
fashion as FILE_SYNC, but with a possible performance drop. If
stable is UNSTABLE, the server is free to commit any part of the
data and the metadata to stable storage, including all or none,
before returning a reply to the client. There is no guarantee
whether or when any uncommitted data will subsequently be committed
to stable storage. The only guarantees made by the server are that
it will not destroy any data without changing the value of verf and
that it will not commit the data and metadata at a level less than
that requested by the client.

The stateid returned from a previous record lock or share
reservation request is provided as part of the argument. The
stateid is used by the server to verify that the associated lock is
still valid and to update lease timeouts for the client.

Upon successful completion, the following results are returned.
The count result is the number of bytes of data written to the
file. The server may write fewer bytes than requested. If so, the
actual number of bytes written starting at location, offset, is
returned.

The server also returns an indication of the level of commitment of
the data and metadata via committed. If the server committed all
data and metadata to stable storage, committed should be set to
FILE_SYNC. If the level of commitment was at least as strong as
DATA_SYNC, then committed should be set to DATA_SYNC. Otherwise,
committed must be returned as UNSTABLE. If stable was FILE_SYNC,
then committed must also be FILE_SYNC: anything else constitutes a
protocol violation. If stable was DATA_SYNC, then committed may be

Expires: December 1999 March 2000 [Page 118] 155]

Draft Protocol Specification NFS version 4 June September 1999

target file

FILE_SYNC or DATA_SYNC: anything else constitutes a protocol
violation. If stable was UNSTABLE, then committed may be either
FILE_SYNC, DATA_SYNC, or UNSTABLE.

The final portion of the result is specified by the current filehandle. write verifier, verf. The offset
specifies
write verifier is a cookie that the offset where client can use to determine
whether the data should server has changed state between a call to WRITE and a
subsequent call to either WRITE or COMMIT. This cookie must be written. An offset
consistent during a single instance of 0 (zero) specifies that the write should start at the beginning NFS version 4 protocol
service and must be unique between instances of the file. The count represents the number of bytes of NFS version 4
protocol server, where uncommitted data that
are to may be written. lost.

If a client writes data to the count is 0 (zero), server with the WRITE will
succeed stable argument set
to UNSTABLE and return the reply yields a count committed response of 0 (zero) subject DATA_SYNC
or UNSTABLE, the client will follow up some time in the future with
a COMMIT operation to permissions
checking. The synchronize outstanding asynchronous data and
metadata with the server's stable storage, barring client error. It
is possible that due to client crash or other error that a
subsequent COMMIT will not be received by the server.

IMPLEMENTATION

It is possible for the server may choose to write fewer bytes than
requested by the client.

Part count bytes of
data. In this case, the server should not return an error unless no
data was written at all. If the server writes less than count
bytes, the client should issue another WRITE to write request the remaining
data.

It is assumed that the act of writing data to a specification file will cause the
time_modified of how the write is file to be performed. The client specifies with the stable parameter updated. However, the
method time_modified
of how the data is to file should not be processed by changed unless the server. If stable
is FILE_SYNC, contents of the server must commit file
are changed. Thus, a WRITE request with count set to 0 should not
cause the time_modified of the data written plus all file
system metadata to be updated.

The definition of stable storage before returning results. has been historically a point of
contention. The following expected properties of stable storage may
help in resolving design issues in the implementation. Stable
storage is persistent storage that survives:

1. Repeated power failures.
2. Hardware failures (of any board, power supply, etc.).
3. Repeated software crashes, including reboot cycle.

This
corresponds to definition does not address failure of the stable storage
module itself.

Expires: March 2000 [Page 156]

Draft Protocol Specification NFS version 2 protocol semantics. Any other
behavior constitutes 4 September 1999

The verifier, is defined to allow a client to detect different
instances of an NFS version 4 protocol violation. If stable is DATA_SYNC,
then the server must commit all of the over which cached,
uncommitted data to stable storage and
enough of may be lost. In the metadata to retrieve most likely case, the data before returning. The verifier
allows the client to detect server implementor reboots. This information is free to implement DATA_SYNC in
required so that the same
fashion as FILE_SYNC, but with a possible performance drop. client can safely determine whether the server
could have lost cached data. If
stable is UNSTABLE, the server is free to commit any part of fails unexpectedly and
the client has uncommitted data and from previous WRITE requests (done
with the metadata to stable storage, including all or none,
before returning a reply argument set to UNSTABLE and in which the client. There is no guarantee
whether or when any uncommitted data will subsequently be result
committed was returned as UNSTABLE as well) it may not have flushed
cached data to stable storage. The only guarantees made by the server are that
it will not destroy any data without changing the value burden of verf recovery is on the
client and
that it the client will not commit need to retransmit the data and metadata at a level less than to the
server.

A suggested verifier would be to use the time that requested by the client.

The stateid returned from a previous record lock server was
booted or share
reservation request is provided as part of the argument. The
stateid is used by time the server was last started (if restarting the
server without a reboot results in lost buffers).

The committed field in the results allows the client to verify that do more
effective caching. If the associated lock server is
still valid and committing all WRITE requests
to update lease timeouts for stable storage, then it should return with committed set to
FILE_SYNC, regardless of the client.

Upon successful completion, value of the following results are returned. stable field in the
arguments. A server that uses an NVRAM accelerator may choose to
implement this policy. The count result is client can use this to increase the number of bytes
effectiveness of the cache by discarding cached data written to that has
already been committed on the
file. The server server.

Some implementations may write fewer bytes than requested. If so, the
actual number return NFS4ERR_NOSPC instead of bytes written starting at location, offset,
NFS4ERR_DQUOT when a user's quota is
returned. exceeded.

ERRORS

NFS4ERR_IO

NFS4ERR_ACCES

NFS4ERR_INVAL

NFS4ERR_FBIG

NFS4ERR_NOSPC

NFS4ERR_ROFS

NFS4ERR_DQUOT

NFS4ERR_STALE

Expires: March 2000 [Page 157]

Draft Protocol Specification NFS version 4 September 1999

NFS4ERR_BADHANDLE

NFS4ERR_SERVERFAULT

NFS4ERR_JUKEBOX

NFS4ERR_LOCKED

NFS4ERR_GRACE

NFS4ERR_FHEXPIRED

NFS4ERR_WRONGSEC

NFS4ERR_MOVED

Expires: March 2000 [Page 158]

Draft Protocol Specification NFS version 4 September 1999

13. NFS Version 4 Callback Procedures

The server also returns an indication of procedures used for callbacks are defined in the level of commitment of following
sections. In the data and metadata via committed. If interest of clarity, the server committed all
data terms "client" and metadata to stable storage, committed should be set
"server" refer to
FILE_SYNC. If NFS clients and servers, despite the level fact that for
an individual callback RPC, the sense of commitment was at least as strong as
DATA_SYNC, then committed should be set to DATA_SYNC. Otherwise,
committed must be returned as UNSTABLE. If stable was FILE_SYNC,
then committed must also be FILE_SYNC: anything else constitutes a
protocol violation. If stable was DATA_SYNC, then committed may these terms would be
precisely the opposite.

13.1. Procedure 0: CB_NULL - No Operation

SYNOPSIS

<null>

ARGUMENT

void;

RESULT

void;

DESCRIPTION

Standard ONCRPC NULL procedure. Void argument, void response.

ERRORS

None.

Expires: December March 2000 [Page 159]

Draft Protocol Specification NFS version 4 September 1999

13.2. Procedure 1: CB_COMPOUND - Compound Operations

SYNOPSIS

compoundargs -> compoundres

ARGUMENT

union cb_opunion switch (unsigned opcode) {
case <OPCODE>: <argument>;
...
};

struct cb_op {
cb_opunion ops;
};

struct CB_COMPOUND4args {
utf8string tag;
cb_op oplist<>;
};

RESULT

union cb_resultdata switch (unsigned resop){
case <OPCODE: <result>;
...
};

struct CB_COMPOUND4res {
nfsstat4 status;
utf8string tag;
cb_resultdata data<>;
};

union opunion switch (unsigned opcode) {
case <OPCODE>: <argument>;
...
};

struct op {
opunion ops;
};

struct COMPOUND4args {

Expires: March 2000 [Page 119] 160]

Draft Protocol Specification NFS version 4 June September 1999

FILE_SYNC or DATA_SYNC: anything else constitutes a protocol
violation. If stable was UNSTABLE, then committed may be either
FILE_SYNC, DATA_SYNC, or UNSTABLE.

The final portion of the result is the write verifier, verf.

utf8string tag;
op oplist<>;
};

DESCRIPTION

The
write verifier CB_COMPOUND procedure is a cookie that the client can use used to determine
whether combine one or more of the server has changed state between
callback procedures into a call to WRITE single RPC request. The main callback
RPC program has two main procedures: CB_NULL and CB_COMPOUND. All
other procedures use the CB_COMPOUND procedure as a
subsequent call to either WRITE or COMMIT. This cookie must be
consistent during a single instance of wrapper.

In the NFS version 4 protocol
service and must be unique between instances processing of the NFS version 4
protocol server, where uncommitted data may be lost.

If a client writes data to CB_COMPOUND procedure, the server with may find
that it does not have the stable argument set available resources to UNSTABLE and the reply yields a committed response of DATA_SYNC execute any or UNSTABLE, all
of the client will follow up some time in procedures within the future with
a COMMIT operation to synchronize outstanding asynchronous data and
metadata with CB_COMPOUND sequence. In this case,
the server's stable storage, barring client error. It
is possible that due to client crash or other error that a
subsequent COMMIT NFS4ERR_RESOURCE will not be received by returned for the server. particular
procedure within the CB_COMPOUND operation where the resource
exhaustion occurred. This assumes that all previous procedures
within the CB_COMPOUND sequence have been evaluated successfully.

IMPLEMENTATION

The CB_COMPOUND procedure is possible for the server used to write fewer than count bytes combine individual procedures
into a single RPC request. The server interprets each of
data. In this case, the server should not return an error unless no
data was written at all.
procedures in turn. If a procedure is executed by the server writes less than count
bytes, the client should issue another WRITE to write the remaining
data.

It is assumed that and
the act status of writing data to a file will cause that procedure is NFS4_OK, then the
time_modified of next procedure in
the file CB_COMPOUND procedure is executed. The server continues this
process until there are no more procedures to be updated. However, executed or one of
the procedures has a status value other than NFS4_OK.

ERRORS

NFS4ERR_RESOURCE

Expires: March 2000 [Page 161]

Draft Protocol Specification NFS version 4 September 1999

13.2.1. Procedure 2: CB_GETATTR - Get Attributes

SYNOPSIS

fh, attrbits -> attrbits, attrvals

ARGUMENT

struct CB_GETATTR4args {
nfs_fh4 fh;
bitmap4 attr_request;
};

RESULT

struct CB_GETATTR4resok {
fattr4 obj_attributes;
};

union CB_GETATTR4res switch (nfsstat4 status) {
case NFS4_OK:
CB_GETATTR4resok resok4;
default:
void;
};

DESCRIPTION

CB_GETATTR is used to obtain the time_modified
of attributes modified by an open
delegate to allow the server to respond to GETATTR requests for a
file should not be changed unless which is the contents subject of an open delegation.

IMPLEMENTATION

The client returns attrbits and the file
are changed. Thus, a WRITE request with count set to 0 should not
cause the time_modified of associated attribute values
only for attributes that it may change (change, time_modify,
object_size). It may further limit the file response to be updated.

The definition of stable storage attributes that
it has been historically a point of
contention. The following expected properties of stable storage may
help in resolving design issues in fact changed during the implementation. Stable
storage is persistent storage that survives:

1. Repeated power failures.
2. Hardware failures (of any board, power supply, etc.).
3. Repeated software crashes, including reboot cycle.

This definition does not address failure scope of the stable storage
module itself. delegation.

Expires: December 1999 March 2000 [Page 120] 162]

Draft Protocol Specification NFS version 4 June September 1999

The verifier, is defined to allow a client to detect different
instances of an

ERRORS

<TBD>

Expires: March 2000 [Page 163]

Draft Protocol Specification NFS version 4 protocol server over which cached,
uncommitted data may be lost. In the most likely case, the verifier
allows the client to detect server reboots. This information September 1999

13.2.2. Procedure 3: CB_RECALL - Recall an Open Delegation

SYNOPSIS

stateid, truncate, fh

ARGUMENT

struct CB_RECALL4args {
stateid4 stateid;
bool truncate;
nfs_fh4 fh;
};

RESULT

struct CB_RECALL4res {
nfsstat4 status;
};

DESCRIPTION

CB_RECALL is
required so that the client can safely determine whether the server
could have lost cached data. If the server fails unexpectedly and
the client has uncommitted data from previous WRITE requests (done
with the stable argument set used to UNSTABLE and in which