draft-ietf-ipatm-ipmc-07.txt  -->   draft-ietf-ipatm-ipmc-08.txt

view Side-By-Side changes


Internet-Draft                                      Grenville Armitage
                                                              Bellcore
                                                  September 20th,
                                                    October 24th, 1995


       Support for Multicast over UNI 3.1 3.0/3.1 based ATM Networks.
                     <draft-ietf-ipatm-ipmc-07.txt>
                     <draft-ietf-ipatm-ipmc-08.txt>


Status of this Memo

   This document was submitted to the IETF IP over ATM WG. Publication
   of this document does not imply acceptance by the IP over ATM WG of
   any ideas expressed within.  Comments should be submitted to the ip-
   atm@matmos.hpl.hp.com mailing list.

   Distribution of this memo is unlimited.

   This memo is an internet draft. 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.  Internet Drafts may be updated, replaced, or obsoleted by
   other documents at any time. It is not appropriate to use Internet
   Drafts as reference material or to cite them other than as a "working
   draft" or "work in progress".

   Please check the lid-abstracts.txt listing contained in the
   internet-drafts shadow directories on ds.internic.net (US East
   Coast), nic.nordu.net (Europe), ftp.isi.edu (US West Coast), or
   munnari.oz.au (Pacific Rim) to learn the current status of any
   Internet Draft.

Abstract

   Mapping the connectionless IP multicast service over the connection
   oriented ATM services provided by UNI 3.1 3.0/3.1 is a non-trivial task.
   This memo describes a mechanism to support the multicast needs of
   Layer 3 protocols in general, and describes its application to IP
   multicasting in particular.

   ATM based IP hosts and routers use a Multicast Address Resolution
   Server (MARS) to support RFC 1112 style Level 2 IP multicast over the
   ATM Forum's UNI 3.1 3.0/3.1 point to multipoint connection service.
   Clusters of endpoints share a MARS and use it to track and
   disseminate information identifying the nodes listed as receivers for given



Armitage                Expires March 20th, April 24th, 1996                 [Page 1]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   given multicast groups. This allows endpoints to establish and manage
   point to multipoint VCs when transmitting to the group.

   The MARS behaviour allows Layer 3 multicasting to be supported using
   either meshes of VCs or ATM level multicast servers. This choice may
   be made on a per-group basis, and is transparent to the endpoints.













































Armitage                Expires March 20th, April 24th, 1996                 [Page 2]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


Contents.

   1. Introduction.
    1.1 The Multicast Address Resolution Server (MARS).
    1.2 The ATM level multicast Cluster.
    1.3 Document overview.
    1.4 Conventions.
   2. The IP multicast service model.
   3. UNI 3.1 3.0/3.1 support for intra-cluster multicasting.
    3.1 VC meshes.
    3.2 Multicast Servers.
    3.3 Tradeoffs.
    3.4 Interaction with local UNI 3.1 3.0/3.1 signalling entity.
   4. Overview of the MARS.
    4.1 Architecture.
    4.2 Control message format.
    4.3 Common Fixed header values fields in MARS control messages.
      4.3.1 Hardware type.
      4.3.2 Protocol type.
      4.3.3 Checksum.
      4.3.4 Extensions Offset.
      4.3.5 Operation code.
      4.3.6 Reserved.
   5. Endpoint (MARS client) interface behaviour.
    5.1 Transmit side behaviour.
      5.1.1 Retrieving Group Membership from the MARS.
      5.1.2 MARS_REQUEST, MARS_MULTI, and MARS_NAK messages.
      5.1.3 Establishing the outgoing multipoint VC.
      5.1.4 Monitoring updates on ClusterControlVC.
        5.1.4.1 Updating the active VCs.
        5.1.4.2 Tracking the Cluster Sequence Number.
      5.1.5 Revalidating a VC's leaf nodes.
        5.1.5.1 When leaf node drops itself.
        5.1.5.2 When a jump is detected in the CSN.
    5.2. Receive side behaviour.
      5.2.1 Format of the MARS_JOIN and MARS_LEAVE Messages.
        5.2.1.1 Important IPv4 default values.
      5.2.2 Retransmission of MARS_JOIN and MARS_LEAVE messages.
      5.2.3 Cluster member registration and deregistration.
    5.3 Support for Layer 3 group management.
    5.4 Support for redundant/backup MARS entities.
      5.4.1 First response to MARS problems.
      5.4.2 Connecting to a backup MARS.
      5.4.3 Dynamic backup lists, and soft redirects.
    5.5 Data path LLC/SNAP encapsulations.
      5.5.1 Type #1 encapsulation.
      5.5.2 Type #2 encapsulation.
      5.5.3 A Type #1 example.



Armitage                Expires April 24th, 1996                 [Page 3]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   6. The MARS in greater detail.
    6.1 Basic interface to Cluster members.
      6.1.1 Response to MARS_REQUEST.
      6.1.2 Response to MARS_JOIN and MARS_LEAVE.
      6.1.3 Generating MARS_REDIRECT_MAP.
      6.1.4 Cluster Sequence Numbers.
    6.2 MARS interface to Multicast Servers (MCSs).
      6.2.1 MARS_REQUESTs for MCS supported groups.
      6.2.2 MARS_MSERV and MARS_UNSERV messages.



Armitage                Expires March 20th, 1996                 [Page 3]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995
      6.2.3 Registering a Multicast Server (MCS).
      6.2.4 Modified response to MARS_JOIN and MARS_LEAVE.
      6.2.5 Sequence numbers for ServerControlVC traffic.
    6.3 Why global sequence numbers?
    6.4 Redundant/Backup MARS Architectures.
   7. How an MCS utilises a MARS.
    7.1 Association with a particular Layer 3 group.
    7.2 Termination of incoming VCs.
    7.3 Management of outgoing VC.
    7.4 Use of a backup MARS.
   8. Support for IP multicast routers.
    8.1 Forwarding into a Cluster.
    8.2 Joining in 'promiscuous' mode.
    8.3 Forwarding across the cluster.
    8.4 Joining in 'semi-promiscous' mode.
    8.5 An alternative to IGMP Queries.
    8.6 CMIs across multiple interfaces.
   9. Multiprotocol applications of the MARS and MARS clients.
   10. Supplementary parameter processing.
    10.1 Interpreting the ar$extoff field.
    10.2 The format of TLVs.
    10.3 Processing MARS messages with TLVs.
    10.4 Initial set of TLV elements.
   11. Key Decisions and open issues.
   Acknowledgments
   References
   Appendix A. Hole punching algorithms.
   Appendix B. Minimising the impact of IGMP in IPv4 environments.
   Appendix C. Further comments on 'Clusters'.
   Appendix D. TLV list parsing algorithm.
   Appendix E. Summary of timer values.
   Appendix F. Pseudo code for MARS operation.










Armitage                Expires March 20th, April 24th, 1996                 [Page 4]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


1.  Introduction.

   Multicasting is the process whereby a source host or protocol entity
   sends a packet to multiple destinations simultaneously using a
   single, local 'transmit' operation. The more familiar cases of
   Unicasting and Broadcasting may be considered to be special cases of
   Multicasting (with the packet delivered to one destination, or 'all'
   destinations, respectively).

   Most network layer models, like the one described in RFC 1112 [1] for
   IP multicasting, assume sources may send their packets to an abstract
   'multicast group addresses'.  Link layer support for such an
   abstraction is assumed to exist, and is provided by technologies such
   as Ethernet.

   ATM is being utilized as a new link layer technology to support a
   variety of protocols, including IP. With RFC 1483 [2] the IETF
   defined a multiprotocol mechanism for encapsulating and transmitting
   packets using AAL5 over ATM Virtual Channels (VCs). However, the ATM
   Forum's currently published signalling specification specifications (UNI 3.0 [8]
   and UNI 3.1 [4]) does not provide the multicast address abstraction.
   Unicast connections are supported by point to point, bidirectional
   VCs. Multicasting is supported through point to multipoint
   unidirectional VCs. The key limitation is that the sender must have
   prior knowledge of each intended recipient, and explicitly establish
   a VC with itself as the root node and the recipients as the leaf
   nodes.

   This document has two broad goals:

      Define a group address registration and membership distribution
      mechanism that allows UNI 3.1 3.0/3.1 based networks to support the
      multicast service of protocols such as IP.

      Define specific endpoint behaviours for managing point to
      multipoint VCs to achieve multicasting of layer 3 packets.

   As the IETF is currently in the forefront of using wide area
   multicasting this document's descriptions will often focus on IP
   service model of RFC 1112.  A final chapter will note the
   multiprotocol application of the architecture.

   This document avoids discussion of one highly non-trivial aspect of
   using ATM - the specification of QoS for VCs being established in
   response to higher layer needs. Research in this area is still very
   formative [7], and so it is assumed that future documents will
   clarify the mapping of QoS requirements to VC establishment. The
   default at this time is that VCs are established with a request for



Armitage                Expires April 24th, 1996                 [Page 5]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   Unspecified Bit Rate (UBR) service, as typified by the IETF's use of



Armitage                Expires March 20th, 1996                 [Page 5]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995
   VCs for unicast IP, described in RFC 1755 [6].

1.1  The Multicast Address Resolution Server (MARS).

   The Multicast Address Resolution Server (MARS) is a superset an extended analog
   of the ATM ARP Server introduced in RFC 1577 [3].  It acts as a
   registry, associating layer 3 multicast group identifiers with the
   ATM interfaces representing the group's members.  MARS messages, based on
   the ATM ARP format, messages
   support the distribution of multicast group membership information
   between MARS and endpoints (hosts or routers).  Endpoint address
   resolution entities query the MARS when a layer 3 address needs to be
   resolved to the set of ATM endpoints making up the group at any one
   time. Endpoints keep the MARS informed when they need to join or
   leave particular layer 3 groups.  To provide for asynchronous
   notification of group membership changes the MARS manages a point to
   multipoint VC out to all endpoints desiring multicast support

   Valid arguments can be made for two different approaches to ATM level
   multicasting of layer 3 packets - through meshes of point to
   multipoint VCs, or ATM level multicast servers (MCS). The MARS
   architecture allows either VC meshes or MCSs to be used on a per-
   group basis.

1.2  The ATM level multicast Cluster.

   Each MARS manages a 'cluster' of ATM-attached endpoints. A Cluster is
   defined as

      The set of ATM interfaces chosen chosing to participate in direct ATM
      connections to achieve multicasting of AAL_SDUs between
      themselves.

   In practice, a Cluster is the set of endpoints that choose to use the
   same MARS to register their memberships and receive their updates
   from.

   By implication of this definition, traffic between interfaces
   belonging to different Clusters passes through an inter-cluster
   device. (In the IP world an inter-cluster device would be an IP
   multicast router with logical interfaces into each Cluster.) This
   document explicitly avoids specifying the nature of inter-cluster
   (layer 3) routing protocols.

   The mapping of clusters to other constrained sets of endpoints (such
   as unicast Logical IP Subnets) is left to each network administrator.
   However, for the purposes of conformance with this document network
   administrators MUST ensure that each Logical IP Subnet (LIS) is



Armitage                Expires March 20th, April 24th, 1996                 [Page 6]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   served by a separate MARS, creating a one-to-one mapping between
   cluster and unicast LIS.  IP multicast routers then interconnect each
   LIS as they do with conventional subnets. (Relaxation of this
   restriction MAY only occur after future research on the interaction
   between existing layer 3 multicast routing protocols and unicast
   subnet boundaries.)

   The term 'Cluster Member' will be used in this document to refer to
   an endpoint that is currently using a MARS for multicast support.
   Thus potential scope of a cluster may be the entire membership of a
   LIS, while the actual scope of a cluster depends on which endpoints
   are actually cluster members at any given time.

1.3  Document overview.

   This document assumes an understanding of concepts explained in
   greater detail in RFC 1112, RFC 1577, UNI 3.1, 3.0/3.1, and RFC 1755 [6].

   Section 2 provides an overview of IP multicast and what RFC 1112
   required from Ethernet.

   Section 3 describes in more detail the multicast support services
   offered by UNI 3.1, 3.0/3.1, and outlines the differences between VC
   meshes and multicast servers (MCSs) as mechanisms for distributing
   packets to multiple destinations.

   Section 4 provides an overview of the MARS and its relationship to
   ATM endpoints. This section also discusses the encapsulation and
   structure of MARS control messages.

   Section 5 substantially defines the entire cluster member endpoint
   behaviour, on both receive and transmit sides. This includes both
   normal operation and error recovery.

   Section 6 summarises the required behaviour of a MARS.

   Section 7 looks at how a multicast server (MCS) interacts with a
   MARS.

   Section 8 discusses how IP multicast routers may make novel use of
   promiscuous and semi-promiscuous group joins. Also discussed is a
   mechanism designed to reduce the amount of IGMP traffic issued by
   routers.

   Section 9 discusses how this document applies in the more general
   (non-IP) case.

   Section 10 summarises the key proposals, and identifies areas for



Armitage                Expires April 24th, 1996                 [Page 7]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   future research that are generated by this MARS architecture.

   The appendices provide discussion on issues that arise out the
   implementation of this document. Appendix A discusses MARS and
   endpoint algorithms for parsing MARS messages. Appendix B describes
   the particular problems introduced by the current IGMP paradigms, and



Armitage                Expires March 20th, 1996                 [Page 7]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995
   possible interim work-arounds.  Appendix C discusses the 'cluster'
   concept in further detail, while Appendix D briefly outlines an
   algorithm for parsing TLV lists.

1.4  Conventions.

   In this document the following coding and packet representation rules
   are used:

      All multi-octet parameters are encoded in big-endian form (i.e.
      the most significant octet comes first).

      In all multi-bit parameters bit numbering begins at 0 for the
      least significant bit when stored in memory (i.e. the n'th bit has
      weight of 2^n).

      A bit that is 'set', 'on', or 'one' holds the value 1.

      A bit that is 'reset', 'off', 'clear', or 'zero' holds the value
      0.


2.  Summary of the IP multicast service model.

   Under IP version 4 (IPv4), addresses in the range of between 224.0.0.0
   and 239.255.255.255 (224.0.0.0/4) are termed 'Class D' or 'multicast
   group' addresses. These abstractly represent all the IP hosts in the
   Internet (or some constrained subset of the Internet) who have
   decided to 'join' the specified group.

   RFC1112 requires that a multicast-capable IP interface must support
   the transmission of IP packets to an IP multicast group address,
   whether or not the node considers itself a 'member' of that group.
   Consequently, group membership is effectively irrelevant to the
   transmit side of the link layer interfaces. When Ethernet is used as
   the link layer (the example used in RFC1112), no address resolution
   is required to transmit packets. An algorithmic mapping from IP
   multicast address to Ethernet multicast address is performed locally
   before the packet is sent out the local interface in the same 'send
   and forget' manner as a unicast IP packet.

   Joining and Leaving an IP multicast group is more explicit on the



Armitage                Expires April 24th, 1996                 [Page 8]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   receive side - with the primitives JoinLocalGroup and LeaveLocalGroup
   affecting what groups the local link layer interface should accept
   packets from. When the IP layer wants to receive packets from a
   group, it issues JoinLocalGroup. When it no longer wants to receive
   packets, it issues LeaveLocalGroup. A key point to note is that
   changing state is a local issue, it has no affect on other hosts



Armitage                Expires March 20th, 1996                 [Page 8]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995
   attached to the Ethernet.

   IGMP is defined in RFC 1112 to support IP multicast routers attached
   to a given subnet. Hosts issue IGMP Report messages when they perform
   a JoinLocalGroup, or in response to an IP multicast router sending an
   IGMP Query. By periodically transmitting queries IP multicast routers
   are able to identify what IP multicast groups have non-zero
   membership on a given subnet.

   A specific IP multicast address, 224.0.0.1, is allocated for the
   transmission of IGMP Query messages. All Host IP multicast hosts must layers issue a
   JoinLocalGroup for 224.0.0.1 during their initialisation. when they intend to participate in IP
   multicasting, and issue a LeaveLocalGroup for 224.0.0.1 when they've
   ceased participating in IP multicasting.

   Each host keeps a list of IP multicast groups it has been
   JoinLocalGroup'd to. When a router issues an IGMP Query on 224.0.0.1
   each host begins to send IGMP Reports for each group it is a member
   of. IGMP Reports are sent to the group address, not 224.0.0.1, "so
   that other members of the same group on the same network can overhear
   the Report" and not bother sending one of their own. IP multicast
   routers conclude that a group has no members on the subnet when IGMP
   Queries no longer elict associated replies.

3. UNI 3.1 3.0/3.1 support for intra-cluster multicasting.

   For the purposes of the MARS protocol, both UNI 3.0 and UNI 3.1
   provide equivalent support for multicasting. Differences between UNI
   3.0 and UNI 3.1 in required signalling elements are covered in RFC
   1755.

   This document will describe its operation in terms of 'generic'
   functions that should be available to clients of a UNI 3.1 3.0/3.1
   signalling entity in a given ATM endpoint. The ATM model broadly
   describes an 'AAL User' as any entity that establishes and manages
   VCs and underlying AAL services to exchange data. An IP over ATM
   interface is a form of 'AAL User' (although the default LLC/SNAP
   encapsulation mode specified in RFC1755 really requires that an 'LLC
   entity' is the AAL User, which in turn supports the IP/ATM
   interface).

   The most fundamental limitations of UNI 3.1's 3.0/3.1's multicast support



Armitage                Expires April 24th, 1996                 [Page 9]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   are:

      Only point to multipoint, unidirectional VCs may be established.

      Only the root (source) node of a given VC may add or remove leaf
      nodes.

   Leaf nodes are identified by their unicast ATM addresses.  UNI 3.1
   3.0/3.1 defines two ATM address formats - native E.164 and NSAP
   (although it must be stressed that the NSAP address is so called
   because it uses the NSAP format - an ATM endpoint is NOT a Network
   layer termination point).  In UNI 3.1 3.0/3.1 an 'ATM Number' is the
   primary identification of an ATM endpoint, and it may use either
   format. Under some circumstances an ATM endpoint must be identified
   by both a native E.164 address (identifying the attachment point of a
   private network to a public network), and an NSAP address ('ATM
   Subaddress')



Armitage                Expires March 20th, 1996                 [Page 9]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995 identifying the final endpoint within the private
   network. For the rest of this document the term will be used to mean
   either a single 'ATM Number' or an 'ATM Number' combined with an 'ATM
   Subaddress'.

3.1 VC meshes.

   The most fundamental approach to intra-cluster multicasting is the
   multicast VC mesh.  Each source establishes its own independent point
   to multipoint VC (a single multicast tree) to the set of leaf nodes
   (destinations) that it has been told are members of the group it
   wishes to send packets to.

   Interfaces that are both senders and group members (leaf nodes) to a
   given group will originate one point to multipoint VC, and terminate
   one VC for every other active sender to the group. This criss-
   crossing of VCs across the ATM network gives rise to the name 'VC
   mesh'.

3.2 Multicast Servers.

   An alternative model has each source establish a VC to an
   intermediate node - the multicast server (MCS). The multicast server
   itself establishes and manages a point to multipoint VC out to the
   actual desired destinations.

   The MCS reassembles AAL_SDUs arriving on all the incoming VCs, and
   then queues them for transmission on its single outgoing point to
   multipoint VC. (Reassembly of incoming AAL_SDUs is required at the
   multicast server as AAL5 does not support cell level multiplexing of
   different AAL_SDUs on a single outgoing VC.)




Armitage                Expires April 24th, 1996                [Page 10]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   The leaf nodes of the multicast server's point to multipoint VC must
   be established prior to packet transmission, and the multicast server
   requires an external mechanism to identify them. A side-effect of
   this method is that ATM interfaces that are both sources and group
   members will receive copies of their own packets back from the MCS
   (An alternative method is for the multicast server to explicitly
   retransmit packets on individual VCs between itself and group
   members. A benefit of this second approach is that the multicast
   server can ensure that sources do not receive copies of their own
   packets.)

   The simplest MCS pays no attention to the contents of each AAL_SDU.
   It is purely an AAL/ATM level device. More complex MCS architectures
   (where a single endpoint serves multiple layer 3 groups) are
   possible, but are beyond the scope of this document. More detailed
   discussion is provided in section 7.




Armitage                Expires March 20th, 1996                [Page 10]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995

3.3 Tradeoffs.

   Arguments over the relative merits of VC meshes and multicast servers
   have raged for some time. Ultimately the choice depends on the
   relative trade-offs a system administrator must make between
   throughput, latency, congestion, and resource consumption. Even
   criteria such as latency can mean different things to different
   people - is it end to end packet time, or the time it takes for a
   group to settle after a membership change? The final choice depends
   on the characteristics of the applications generating the multicast
   traffic.

   If we focussed on the data path we might prefer the VC mesh because
   it lacks the obvious single congestion point of an MCS.  Throughput
   is likely to be higher, and end to end latency lower, because the
   mesh lacks the intermediate AAL_SDU reassembly that must occur in
   MCSs. The underlying ATM signalling system also has greater
   opportunity to ensure optimal branching points at ATM switches along
   the multicast trees originating on each source.

   However, resource consumption will be higher. Every group member's
   ATM interface must terminate a VC per sender (consuming on-board
   memory for state information, instance of an AAL service, and
   buffering in accordance with the vendors particular architecture). On
   the contrary, with a multicast server only 2 VCs (one out, one in)
   are required, independent of the number of senders. The allocation of
   VC related resources is also lower within the ATM cloud when using a
   multicast server. These points may be considered to have merit in
   environments where VCs across the UNI or within the ATM cloud are
   valuable (e.g. the ATM provider charges on a per VC basis), or AAL
   contexts are limited in the ATM interfaces of endpoints.



Armitage                Expires April 24th, 1996                [Page 11]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   If we focus on the signalling load then MCSs have the advantage when
   faced with dynamic sets of receivers. Every time the membership of a
   multicast group changes (a leaf node needs to be added or dropped),
   only a single point to multipoint VC needs to be modified when using
   an MCS. This generates a single signalling event across the MCS's
   UNI. However, when membership change occurs in a VC mesh, signalling
   events occur at the UNIs of every traffic source - the transient
   signalling load scales with the number of sources. This has obvious
   ramifications if you define latency as the time for a group's
   connectivity to stabilise after change (especially as the number of
   senders increases).

   Finally, as noted above, MCSs introduce a 'reflected packet' problem,
   which requires additional per-AAL_SDU information to be carried in
   order for layer 3 sources to detect their own AAL_SDUs coming back.




Armitage                Expires March 20th, 1996                [Page 11]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995

   The MARS architecture allows system administrators to utilize either
   approach on a group by group basis.

3.4 Interaction with local UNI 3.1 3.0/3.1 signalling entity.

   The following generic signalling functions are presumed to be
   available to local AAL Users:

   L_CALL_RQ     - Establish a unicast VC to a specific endpoint.
   L_MULTI_RQ    - Establish multicast VC to a specific endpoint.
   L_MULTI_ADD   - Add new leaf node to previously established VC.
   L_MULTI_DROP  - Remove specific leaf node from established VC.
   L_RELEASE     - Release unicast VC, or all Leaves of a multicast VC.

   The signalling exchanges and local information passed between AAL
   User and UNI 3.1 3.0/3.1 signalling entity with these functions are
   outside the scope of this document.

   The following indications are assumed to be available to AAL Users,
   generated by by the local UNI 3.1 3.0/3.1 signalling entity:

   L_ACK          - Succesful completion of a local request.
   L_REMOTE_CALL  - A new VC has been established to the AAL User.
   ERR_L_RQFAILED - A remote ATM endpoint rejected an L_CALL_RQ,
                    L_MULTI_RQ, or L_MULTI_ADD.
   ERR_L_DROP     - A remote ATM endpoint dropped off an existing VC.
   ERR_L_RELEASE  - An existing VC was terminated.

   The signalling exchanges and local information passed between AAL
   User and UNI 3.1 3.0/3.1 signalling entity with these functions are
   outside the scope of this document.




Armitage                Expires April 24th, 1996                [Page 12]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


4.  Overview of the MARS.

   The MARS may reside within any ATM endpoint that is directly
   addressable by the endpoints it is serving. Endpoints wishing to join
   a multicast cluster must be configured with the ATM address of the
   node on which the cluster's MARS resides.  (Section 5.4 describes how
   backup MARSs may be added to support the activities of a cluster.
   References to 'the MARS' in following sections will be assumed to
   mean the acting MARS for the cluster.)

4.1  Architecture.

   Architecturally the MARS is an evolution of the RFC 1577 ARP Server.
   Whilst the ARP Server keeps a table of {IP,ATM} address pairs for all
   IP endpoints in an LIS, the MARS keeps extended tables of {layer 3



Armitage                Expires March 20th, 1996                [Page 12]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995
   address, ATM.1, ATM.2, ..... ATM.n} mappings. It can either be
   configured with certain mappings, or dynamically 'learn' mappings.
   The format of the {layer 3 address} field is generally not
   interpreted by the MARS (except for a few special cases, described
   later).

   A single ATM node may support multiple logical MARSs, each of which
   support a separate cluster. The restriction is that each MARS has a
   unique ATM address (e.g. a different SEL field in the NSAP address of
   the node on which the multiple MARSs reside).  By definition a single
   instance of a MARS may not support more than one cluster.

   The MARS distributes group membership update information to cluster
   members over a point to multipoint VC known as the ClusterControlVC.
   Additionally, when Multicast Servers (MCSs) are being used it also
   establishes a separate point to multipoint VC out to registered MCSs,
   known as the ServerControlVC.  All cluster members are leaf nodes of
   ClusterControlVC. All registered multicast servers are leaf nodes of
   ServerControlVC (described further in section 6).

   The MARS does NOT take part in the actual multicasting of layer 3
   data packets.

4.2  Control message format.

   The MARS message format is an extension of the ATM ARP message
   format.

   By default all MARS control messages MUST be LLC/SNAP encapsulated
   in
   using the same manner as ATMARP messages:

      [0xAA-AA-03][0x00-00-00][0x08-06][MARS following codepoints:

      [0xAA-AA-03][0x00-00-5E][0x00-03][MARS control message]
          (LLC)       (OUI)     (PID)

   (This is a PID from the IANA OUI.)




Armitage                Expires April 24th, 1996                [Page 13]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   MARS control messages may further be subdivided.

      [MARS header][Layer 3 and/or ATM addresses][Supplementary are made up of 4 major components:

      [Fixed header][Mandatory fields][Addresses][Supplementary TLVs]

   [MARS

   [Fixed header] contains fields indicating the operation being
   performed and the layer 3 protocol being referred to (e.g IPv4, IPv6,
   AppleTalk, etc).  (The choice of common encapsulation and The fixed header
   format means that MARS also carries checksum information,
   and ARP Server functionality may hooks to allow this basic control message structure to be
   implemented within a common entity, and share a client-server VC, if
   the implementor so chooses.) re-used
   by other query/response protocols.

   The format of the following [Layer 3 and/or ATM addresses] area in
   the MARS message depends [Mandatory fields] section carries fixed width parameters that
   depend on the operation type indicated in the [MARS [Fixed header].

   The following [Addresses] area carries variable length fields for
   source and target addresses - both hardware (e.g. ATM) and layer 3
   (e.g. IPv4). These provide the fundamental information that the
   registrations, queries, and updates use and operate on.




Armitage                Expires March 20th, 1996                [Page 13]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


   A For the MARS
   protocol fields in [Fixed header] indicate how to interpret the
   contents of [Addresses].

   [Supplementary TLVs] represents an optional list of TLV (type,
   length, value) encoded information elements that may be appended to
   provide supplementary information.  This feature is described in
   further detail in section 10.

   MARS messages contain variable length address fields. In all cases
   null addresses MUST SHALL be encoded as zero length, and have no space
   allocated in the message.

4.3  Common header values in MARS messages.

   The [MARS header] specifically consists

   (Unique LLC/SNAP encapsulation of MARS control messages means MARS
   and ARP Server functionality may be implemented within a common
   entity, and share a client-server VC, if the following fields:

      Data:
       ar$hrd     16 implementor so chooses.
   Note that the LLC/SNAP codepoint for MARS is different to the
   codepoint used for ATMARP.)

4.3  Fixed header fields in MARS control messages.

   The [Fixed header] has the following format:

      Data:
       ar$hrd      16 bits  Hardware type ( 19 decimal, 0x13 hex)
       ar$pro type.
       ar$pro.type 16 bits  Protocol type type.
       ar$pro.snap 40 bits  Optional SNAP extension to protocol type.
       ar$hdrrsv   24 bits  Reserved. Unused by MARS control protocol.
       ar$chksum   16 bits  Checksum across entire MARS message.
       ar$extoff   16 bits  Extensions Offset.
       ar$op       16 bits  Operation code.
       ar$shtl      8 bits  Type & length of source ATM number number.



Armitage                Expires April 24th, 1996                [Page 14]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


       ar$sstl      8 bits  Type & length of source ATM subaddress
       ar$op      16 bits  Operation code.
       ar$extoff  32 bits  Extensions Offset.
              [...rest of subaddress.

   ar$shtl and ar$sstl provide information regarding the source's
   hardware (ATM) address. In the MARS protocol these fields are always
   present, as every MARS message....] message carries a non-null source ATM address.
   In all cases the ar$hrd field is set to 0x13, to indicate source ATM as address is the
   underlying hardware type.

   The ar$op first variable length
   field identifies whether in the message is a MARS, ARP, or
   Inverse ARP message. [Addresses] section.

   The values used by this document other fields in [Fixed header] are summarised described in section 11.

   The ar$pro field identifies the higher layer protocol whose following
   subsections.

4.3.1  Hardware type.

   ar$hrd defines the type of 'hardware' addresses
   are being carried and or mapped to hardware carried.  When
   ar$hrd = 0x13 the addresses identified by are ATM addresses. Interpretation of the
   ATM number and subaddress fields when ar$hrd field. If the ar$op != 0x13 is for future
   definition. The remainder of this document assumes that ar$hrd =
   0x13.

4.3.2  Protocol type.

   The ar$pro.type field indicates a MARS message, the
   ar$pro value represents is a protocol from 16 bit unsigned integer representing the
   following two sets: number space:

      0x0000 to 0x05FF 0x00FF  Protocols defined by the equivalent NPLIDs.
      0x0100 to 0x03FF  Reserved for future use by the IETF.
      0x0400 to 0x04FF  Allocated for use by the ATM Forum.
      0x0500 to 0x05FF  Experimental/Local use.
      0x0600 to 0xFFFF  Protocols defined by the equivalent Ethertypes.

   The use of

   (based on the ar$pro field observations that valid Ethertypes are never smaller
   than 0x600, and NPLIDs never larger than 0xFF.)

   The NLPID value of 0x80 is described further used to indicate a SNAP encoded extension
   is being used to encode the protocol type. When ar$pro.type == 0x80
   the SNAP extension is encoded in section 9. the ar$pro.snap field.  This is
   termed the 'long form' protocol ID.

   If ar$pro != 0x80 then the ar$pro.snap field MUST be zero on transmit
   and ignored on receive. The ar$extoff ar$pro.type field itself identifies the existence and location of
   supplementary parameters. Its use
   protocol being referred to. This is described termed the 'short form' protocol
   ID.

   In all cases, where a protocol has an assigned number in section 10.

   The remaining fields are the
   ar$pro.type space (excluding 0x80) the short form MUST be used in manners specific when
   transmitting MARS messages. Additionally, where a protocol has valid
   short and long forms of identification, receivers MAY choose to
   recognise the operation
   being performed, and are described in later sections. long form.



Armitage                Expires March 20th, April 24th, 1996                [Page 14] 15]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


5.  Endpoint (MARS client) interface behaviour.

   An endpoint is best thought


   ar$pro.type values other than 0x80 MAY have 'long forms' defined in
   future documents.

   For the remainder of as a 'shim' this document references to ar$pro SHALL be
   interpreted to mean ar$pro.type, or 'convergence' layer,
   sitting between a layer 3 protocol's link layer interface and the
   underlying UNI 3.1 service. An endpoint ar$pro.type in this context can exist combination with
   ar$pro.snap as appropriate.

   The use of different protocol types is described further in section
   9.


4.3.3 Checksum.

   The ar$chksum field carries a host or a router - any entity that requires a generic 'layer 3 over
   ATM' interface to support layer 3 multicast.  It is broken into two
   key subsections - one for standard IP checksum calculated across
   the transmit side, and one for entire MARS control message (excluding the receive
   side.

   Multiple logical ATM interfaces may be supported by a single physical
   ATM interface (for example, using different SEL values in the NSAP
   formatted address assigned LLC/SNAP header). The
   field is set to zero before performing the physical ATM interface). Therefore
   implementors MUST allow for multiple independent 'layer 3 over ATM'
   interfaces too, each with its own configured MARS (or table of MARSs,
   as discussed in section 5.4), and ability to be attached to checksum calculation.

   As the same
   or different clusters.

   The initial signalling path between a MARS client (managing an
   endpoint) and its associated entire LLC/SNAP encapsulated MARS message is a transient point to point,
   bidirectional VC.  This VC is established protected by the MARS client, and is
   used
   32 bit CRC of the AAL5 transport, implementors MAY choose to send queries to, and receive replies from, ignore
   the MARS. It has
   an associated idle timer, and is dismantled if not used for a
   configurable period of time. The minimum suggested value for this
   time checksum facility. If no checksum is 1 minute, and the RECOMMENDED default calculated these bits MUST
   be reset before transmission. If no checksum is 20 minutes.  (Where
   the MARS and ARP Server are co-resident, performed on
   reception, this VC may field MUST be used for both
   ATM ARP traffic and MARS control traffic.)

   The remaining signalling path is ClusterControlVC, to which the MARS
   client ignored. If a receiver is added as capable of
   validating a leaf node when checksum it registers (described in
   section 5.2.3).

   Most of this specification MUST only perform the validation when the
   received ar$chksum field is concerned non-zero. Messages arriving with managing and
   distributing information that allows the establishment
   ar$chksum of VCs for
   actually carrying layer 3 data packets. 0 are always considered valid.

4.3.4 Extensions Offset.

   The actual format of the data
   carried on these VCs is almost completely outside ar$extoff field identifies the scope existence and location of this
   specification.  However, when using MCSs (described in section 3)
   endpoints need to filter out the reflected packets that can occur.
   The solution to this problem in a general way requires the an
   optional supplementary parameters list. Its use of
   additional per-packet encapsulation. This is discussed described in
   section 5.5

5.1  Transmit side behaviour. 10.

4.3.5 Operation code.

   The following description will often be in terms of an IPv4/ATM
   interface that ar$op field is capable further subdivided into two 8 bit fields -
   ar$op.version (leading octet) and ar$op.type (trailing octet).
   Together they indicate the nature of transmitting packets to a Class D
   address at any time, without prior warning. It the control message, and the
   context within which its [Mandatory fields], [Addresses], and
   [Supplementary TLVs] should be trivial for
   an implementor to generalise interpreted.

      ar$op.version
         0               MARS protocol defined in this behaviour to document.
         0x01 - 0xEF     Reserved for future use by the requirements IETF.
         0xF0 - 0xFE     Allocated for use by the ATM Forum.
         0xFF            Experimental/Local use.

      ar$op.type
         Value indicates operation being performed, within context of
   another layer 3 data protocol.



Armitage                Expires March 20th, April 24th, 1996                [Page 15] 16]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   When a packet arrives for transmission, and there is no outgoing VC
   already marked as serving the packet's multicast destination address,


         the MARS is queried for control protocol version indicated by ar$op.version.

   For the set rest of ATM endpoints currently making up this document references to the multicast group. ar$op value SHALL be
   taken to mean ar$op.type, with ar$op.version = 0x00. The query values used
   in this document are summarised in section 11.

   (Note this number space is executed by issuing a MARS_REQUEST.  The reply from independent of the
   MARS ATMARP operation code
   number space.)

4.3.6 Reserved.

   ar$hdrrsv may take one of two forms:

      MARS_MULTI - Sequence of MARS_MULTI messages returning the set of
                   ATM endpoints that are to be leaf nodes of the
                   outgoing VC.

      MARS_NAK - No mapping found, group subdivided and assigned specific meanings for other
   control protocols indicated by ar$op.version != 0.


5.  Endpoint (MARS client) interface behaviour.

   An endpoint is empty.

   The formats best thought of these messages are described in section 5.1.2.

   Outgoing VCs are established with a request for Unspecified Bit Rate
   (UBR) service, as typified by a 'shim' or 'convergence' layer,
   sitting between a layer 3 protocol's link layer interface and the IETF's use of VCs for unicast IP,
   described
   underlying UNI 3.0/3.1 service. An endpoint in RFC 1755 [6].  Future documents may vary this approach
   and allow the specification of different ATM traffic parameters from
   locally configured information context can exist
   in a host or parameters obtained through some
   external means.

5.1.1   Retrieving Group Membership from the MARS.

   If the MARS had no mapping a router - any entity that requires a generic 'layer 3
   over ATM' interface to support layer 3 multicast.  It is broken into
   two key subsections - one for the desired Class D address a MARS_NAK
   will be returned. In this case transmit side, and one for the IP packet MUST
   receive side.

   Multiple logical ATM interfaces may be discarded
   silently. If supported by a match is found single physical
   ATM interface (for example, using different SEL values in the MARS's tables it proceeds NSAP
   formatted address assigned to
   return addresses ATM.1 through ATM.n in a sequence of one or more
   MARS_MULTIs.  A simple mechanism is used to detect and recover from
   loss of MARS_MULTI messages.

   (If the client learns that there is no other group member physical ATM interface). Therefore
   implementors MUST allow for multiple independent 'layer 3 over ATM'
   interfaces too, each with its own configured MARS (or table of MARSs,
   as discussed in section 5.4), and ability to be attached to the
   cluster - the MARS returns a MARS_NAK same
   or returns different clusters.

   The initial signalling path between a MARS_MULTI with
   the MARS client as the only member - it MUST delay sending out a new
   MARS_REQUEST for that group for a period no less than 5 seconds (managing an
   endpoint) and
   no more than 10 seconds.)

   Each MARS_MULTI carries its associated MARS is a boolean field x, transient point to point,
   bidirectional VC.  This VC is established by the MARS client, and a 15 bit integer field
   y - expressed as MARS_MULTI(x,y). Field y acts as a sequence number,
   starting at 1 is
   used to send queries to, and incrementing for each MARS_MULTI sent.  Field x
   acts as receive replies from, the MARS. It has
   an 'end associated idle timer, and is dismantled if not used for a
   configurable period of reply' marker. When x == time. The minimum suggested value for this
   time is 1 minute, and the MARS response RECOMMENDED default is
   considered complete.

   In addition, each MARS_MULTI 20 minutes.  (Where
   the MARS and ARP Server are co-resident, this VC may carry multiple be used for both
   ATM addresses from ARP traffic and MARS control traffic.)

   The remaining signalling path is ClusterControlVC, to which the set {ATM.1, ATM.2, .... ATM.n}. A MARS MUST minimise the number
   of MARS_MULTIs transmitted by placing
   client is added as many group member's
   addresses in a single MARS_MULTI as possible. leaf node when it registers (described in
   section 5.2.3).

   The limit on majority of this document covers the length distribution of information



Armitage                Expires March 20th, April 24th, 1996                [Page 16] 17]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   of an individual MARS_MULTI message MUST be


   allowing endpoints to establish and manage outgoing point to
   multipoint VCs - the MTU forwarding paths for multicast traffic to
   particular multicast groups. The actual format of the underlying
   VC.

   For example, assume n ATM addresses must be returned, each MARS_MULTI AAL_SDUs sent
   on these VCs is limited almost completely outside the scope of this
   specification.  However, endpoints are not expected to know whether
   their forwarding path leads directly to only p ATM addresses, and p << n. This would require a
   sequence of k MARS_MULTI messages (where k = (n/p)+1, using integer
   arithmetic), transmitted as follows:

      MARS_MULTI(0,1) carries back {ATM.1 ... ATM.p}
      MARS_MULTI(0,2) carries back {ATM.(p+1) ... ATM.(2p)}
            [.......]
      MARS_MULTI(1,k) carries back { ... ATM.n}

   If k == 1 then only MARS_MULTI(1,1) is sent.

   Typical failure mode will be losing one multicast group's members
   or more of MARS_MULTI(0,1)
   through MARS_MULTI(0,k-1). to an MCS (described in section 3). This is detected when y jumps by more than
   one between consecutive MARS_MULTI's. An alternative failure mode is
   losing MARS_MULTI(1,k).  A timer MUST be implemented requires additional per-
   packet encapsulation (described in section 5.5) to flag aid in the
   failure of the last MARS_MULTI to arrive. A default value
   detection of 10
   seconds is RECOMMENDED.

   If a 'sequence jump' reflected AAL_SDUs.

5.1  Transmit side behaviour.

   The following description will often be in terms of an IPv4/ATM
   interface that is detected, the host MUST wait capable of transmitting packets to a Class D
   address at any time, without prior warning. It should be trivial for
   an implementor to generalise this behaviour to the
   MARS_MULTI(1,k), discard all results, and repeat the MARS_REQUEST.

   If requirements of
   another layer 3 data protocol.

   When a timeout occurs, the host MUST discard all results, and repeat local Layer 3 entity passes down a packet for transmission,
   the MARS_REQUEST.

   (Corruption of cell contents will lead endpoint first ascertains whether an outbound path to loss of a MARS_MULTI
   through AAL5 CPCS_PDU reassembly failure, which will be detected
   through the mechanisms described above.)
   destination multicast group already exists. If it does not, the MARS
   is managing queried for a cluster set of ATM endpoints spread across
   different but directly accessible that represent an appropriate
   forwarding path. (The ATM networks it will not be able to
   return all endpoints may represent the actual group
   members in within the cluster, or a single MARS_MULTI. set of one or more MCSs. The MARS_MULTI
   message format allows for
   endpoint does not distinguish between either E.164, ISO NSAP, or (E.164 + NSAP) case. Section 6.2
   describes the MARS behaviour that leads to be returned MCSs being supplied as ATM addresses. However, each MARS_MULTI message may
   only return ATM addresses of the same type and length.
   forwarding path for a multicast group.)

   The returned
   addresses MUST be grouped according to type (E.164, ISO NSAP, or
   both) and returned in query is executed by issuing a sequence MARS_REQUEST.  The reply from the
   MARS may take one of separate two forms:

      MARS_MULTI parts.

5.1.2   MARS_REQUEST, MARS_MULTI, and MARS_NAK messages.

   MARS_REQUEST is shown below. It is indicated an 'operation type
   value' (ar$op) - Sequence of 11.

   The multicast address being resolved is placed into MARS_MULTI messages returning the set of
                   ATM endpoints that are to be leaf nodes of an
                   outgoing point to multipoint VC (the forwarding
                   path).

      MARS_NAK - No mapping found, group is empty.

   The formats of these messages are described in section 5.1.2.

   Outgoing VCs are established with a request for Unspecified Bit Rate
   (UBR) service, as typified by the target
   protocol address field (ar$tpa), IETF's use of VCs for unicast IP,
   described in RFC 1755 [6].  Future documents may vary this approach
   and allow the target hardware address is specification of different ATM traffic parameters from
   locally configured information or parameters obtained through some
   external means.





Armitage                Expires March 20th, April 24th, 1996                [Page 17] 18]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   set to null (ar$thtl and ar$tstl both zero).

   In IPv4 environments


5.1.1   Retrieving Group Membership from the protocol type (ar$pro) is 0x800 and MARS.

   If the
   target protocol MARS had no mapping for the desired Class D address length (ar$tpln) MUST a MARS_NAK
   will be set to 4. The source
   fields returned. In this case the IP packet MUST contain be discarded
   silently. If a match is found in the ATM number MARS's tables it proceeds to
   return addresses ATM.1 through ATM.n in a sequence of one or more
   MARS_MULTIs.  A simple mechanism is used to detect and subaddress recover from
   loss of MARS_MULTI messages.

   (If the client
   issuing learns that there is no other group member in the
   cluster - the MARS returns a MARS_NAK or returns a MARS_MULTI with
   the client as the only member - it MUST delay sending out a new
   MARS_REQUEST (the subaddress MAY be null).

      Data:
       ar$hrd     16 bits  Hardware type ( 19 decimal, 0x13 hex)
       ar$pro     16 bits  Protocol type
       ar$shtl     8 bits  Type & length for that group for a period no less than 5 seconds and
   no more than 10 seconds.)

   Each MARS_MULTI carries a boolean field x, and a 15 bit integer field
   y - expressed as MARS_MULTI(x,y). Field y acts as a sequence number,
   starting at 1 and incrementing for each MARS_MULTI sent.  Field x
   acts as an 'end of source reply' marker. When x == 1 the MARS response is
   considered complete.

   In addition, each MARS_MULTI may carry multiple ATM addresses from
   the set {ATM.1, ATM.2, .... ATM.n}. A MARS MUST minimise the number (q)
       ar$sstl     8 bits  Type & length of source ATM subaddress (r)
       ar$op      16 bits  Operation code (MARS_REQUEST)
       ar$extoff  32 bits  Extensions Offset.
       ar$spln     8 bits  Length
   of source protocol address (s)
       ar$thtl     8 bits  Type & length of target ATM number (x)
       ar$tstl     8 bits  Type & MARS_MULTIs transmitted by placing as many group member's
   addresses in a single MARS_MULTI as possible. The limit on the length
   of target ATM subaddress (y)
       ar$tpln     8 bits  Length an individual MARS_MULTI message MUST be the MTU of target multicast group address (z)
       ar$sha     qoctets  source ATM number
       ar$ssa     roctets  source ATM subaddress
       ar$spa     soctets  source protocol address
       ar$tha     xoctets  target the underlying
   VC.

   For example, assume n ATM number
       ar$tsa     yoctets  target addresses must be returned, each MARS_MULTI
   is limited to only p ATM subaddress
       ar$tpa     zoctets  target multicast group address

   Following the RFC1577 approach, the ar$shtl, ar$sstl, ar$thtl addresses, and
   ar$tstl fields are coded as follows:

                7 6 5 4 3 2 p << n. This would require a
   sequence of k MARS_MULTI messages (where k = (n/p)+1, using integer
   arithmetic), transmitted as follows:

      MARS_MULTI(0,1) carries back {ATM.1 ... ATM.p}
      MARS_MULTI(0,2) carries back {ATM.(p+1) ... ATM.(2p)}
            [.......]
      MARS_MULTI(1,k) carries back { ... ATM.n}

   If k == 1 0
               +-+-+-+-+-+-+-+-+
               |0|x|  length   |
               +-+-+-+-+-+-+-+-+

   The most significant bit then only MARS_MULTI(1,1) is reserved and sent.

   Typical failure mode will be losing one or more of MARS_MULTI(0,1)
   through MARS_MULTI(0,k-1). This is detected when y jumps by more than
   one between consecutive MARS_MULTI's. An alternative failure mode is
   losing MARS_MULTI(1,k).  A timer MUST be set implemented to zero.  The
   second most significant bit (x) is a flag indicating whether the ATM
   address being referred
   failure of the last MARS_MULTI to is in:

      - ATM Forum NSAPA format (x = 0).
      - Native E.164 format (x = 1).

   The bottom 6 bits is an unsigned integer arrive. A default value indicating the length of 10
   seconds is RECOMMENDED.

   If a 'sequence jump' is detected, the associated ATM address in octets.

   The ar$spln and ar$tpln fields are unsigned 8 bit integers, giving
   the length in octets of host MUST wait for the source and target protocol address fields
   respectively.

   MARS packets use true variable length fields. A null (non-existant)



Armitage                Expires March 20th, April 24th, 1996                [Page 18] 19]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   address MUST be coded as zero length,


   MARS_MULTI(1,k), discard all results, and no space allocated for it
   in repeat the message body.

   MARS_NAK is MARS_REQUEST.

   If a timeout occurs, the MARS_REQUEST returned with operation type value of 16
   (decimal). All other fields are left unchanged from host MUST discard all results, and repeat
   the MARS_REQUEST
   (e.g. do not transpose MARS_REQUEST.

   A final failure mode involves the source MARS Sequence Number (described in
   section 5.1.4.2 and target information. In carried in each part of a multi-part MARS_MULTI).
   If its value changes during the reception of a multi-part MARS_MULTI
   the host MUST wait for the MARS_MULTI(1,k), discard all
   cases MARS clients use results, and
   repeat the source address fields MARS_REQUEST.

   (Corruption of cell contents will lead to identify their
   own messages coming back).

   The loss of a MARS_MULTI message
   through AAL5 CPCS_PDU reassembly failure, which will be detected
   through the mechanisms described above.)

   If the MARS is identified by an ar$op value managing a cluster of 12. endpoints spread across
   different but directly accessible ATM networks it will not be able to
   return all the group members in a single MARS_MULTI. The MARS_MULTI
   message format is:

      Data:
       ar$hrd     16 bits  Hardware allows for either E.164, ISO NSAP, or (E.164 + NSAP)
   to be returned as ATM addresses. However, each MARS_MULTI message may
   only return ATM addresses of the same type ( 19 decimal, 0x13 hex)
       ar$pro     16 bits  Protocol and length. The returned
   addresses MUST be grouped according to type
       ar$shtl     8 bits  Type & length of source ATM number (q)
       ar$sstl     8 bits  Type & length (E.164, ISO NSAP, or
   both) and returned in a sequence of source ATM subaddress (r)
       ar$op      16 bits  Operation code (MARS_MULTI)
       ar$extoff  32 bits  Extensions Offset.
       ar$spln     8 bits  Length separate MARS_MULTI parts.

5.1.2   MARS_REQUEST, MARS_MULTI, and MARS_NAK messages.

   MARS_REQUEST is shown below. It is indicated an 'operation type
   value' (ar$op) of source 1.

   The multicast address being resolved is placed into the the target
   protocol address (s)
       ar$thtl     8 bits  Type & field (ar$tpa), and the target hardware address is
   set to null (ar$thtl and ar$tstl both zero).

   In IPv4 environments the protocol type (ar$pro) is 0x800 and the
   target protocol address length (ar$tpln) MUST be set to 4. The source
   fields MUST contain the ATM number and subaddress of the client
   issuing the MARS_REQUEST (the subaddress MAY be null).

      Data:
       ar$hrd      16 bits  Hardware type.
       ar$pro.type 16 bits  Protocol type.
       ar$pro.snap 40 bits  Optional SNAP extension to protocol type.
       ar$hdrrsv   24 bits  Reserved. Unused by MARS control protocol.
       ar$chksum   16 bits  Checksum across entire MARS message.
       ar$extoff   16 bits  Extensions Offset.
       ar$op       16 bits  Operation code (MARS_REQUEST = 1)
       ar$shtl      8 bits  Type & length of source ATM number.
       ar$sstl      8 bits  Type & length of source ATM subaddress.



Armitage                Expires April 24th, 1996                [Page 20]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


       ar$spln      8 bits  Length of source protocol address (s)
       ar$thtl      8 bits  Type & length of target ATM number (x)
       ar$tstl      8 bits  Type & length of target ATM subaddress (y)
       ar$tpln      8 bits  Length of target multicast group address (z)
       ar$tnum    16 bits  Number of target ATM addresses returned (N).
       ar$seqxy   16 bits  Boolean flag x and sequence number y.
       ar$msn     32
       ar$pad      64 bits  MARS Sequence Number.  Padding (aligns ar$sha with MARS_MULTI).
       ar$sha      qoctets  source ATM number
       ar$ssa      roctets  source ATM subaddress
       ar$spa      soctets  source protocol address
       ar$tpa     zoctets  target multicast group address
       ar$tha.1   xoctets  target ATM number 1
       ar$tsa.1   yoctets  target ATM subaddress 1
       ar$tha.2
       ar$tha      xoctets  target ATM number 2
       ar$tsa.2
       ar$tsa      yoctets  target ATM subaddress 2
                 [.......]
       ar$tha.N   xoctets  target ATM number N
       ar$tsa.N   yoctets
       ar$tpa      zoctets  target ATM subaddress N

   The source protocol and ATM multicast group address fields are copied directly from
   the MARS_REQUEST that this MARS_MULTI is in response to (not

   Following the MARS
   itself).

   ar$seqxy is coded with flag x in RFC1577 approach, the leading bit, ar$shtl, ar$sstl, ar$thtl and sequence number
   y
   ar$tstl fields are coded as an unsigned integer in the remaining 15 bits.

          |  1st octet    |   2nd octet   |
           7 6 5 4 3 2 1 0 follows:

                7 6 5 4 3 2 1 0
          +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+



Armitage                Expires March 20th, 1996                [Page 19]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


          |x|                 y
               +-+-+-+-+-+-+-+-+
               |0|x|  length   |
          +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   ar$tnum
               +-+-+-+-+-+-+-+-+

   The most significant bit is reserved and MUST be set to zero.  The
   second most significant bit (x) is a flag indicating whether the ATM
   address being referred to is in:

      - ATM Forum NSAPA format (x = 0).
      - Native E.164 format (x = 1).

   The bottom 6 bits is an unsigned integer value indicating how many pairs the length
   of
   {ar$tha,ar$tsa} (i.e. how many group member's the associated ATM addresses) are
   present address in octets. If this value is zero the message. ar$msn
   flag x is an ignored.

   The ar$spln and ar$tpln fields are unsigned 32 8 bit number filled integers, giving
   the length in
   by octets of the source and target protocol address fields
   respectively.

   MARS before transmitting each MARS_MULTI. Its packets use is described
   further in section 5.1.4.

   As an example, assume we have a multicast cluster using 4 byte
   protocol addresses, 20 byte ATM numbers, true variable length fields. A null (non-existant)
   address MUST be coded as zero length, and 0 byte ATM subaddresses.
   For n group members in a single MARS_MULTI we require a (48 + 20n)
   byte message. If we assume the default MTU of 9180 bytes, we can
   return a maximum of 456 group member's addresses no space allocated for it
   in a single
   MARS_MULTI.

5.1.3   Establishing the outgoing multipoint VC.

   Following message body.

   MARS_NAK is the completion MARS_REQUEST returned with operation type value of 6.
   All other fields are left unchanged from the MARS_MULTI reply MARS_REQUEST (e.g. do
   not transpose the endpoint may
   establish a new point source and target information. In all cases MARS
   clients use the source address fields to multipoint VC, or reuse an existing one.

   If establishing a new VC, an L_MULTI_RQ identify their own messages
   coming back).

   The MARS_MULTI message is issued for ATM.1, followed identified by an L_MULTI_ADD for every member of the set {ATM.2, ....ATM.n}
   (assuming the set is non-null). The packet is then transmitted over
   the newly created VC just as it would be for a unicast VC.

   After transmitting the packet, the local interface holds the VC open
   and marks it as the active path out ar$op value of the host for any subsequent IP
   packets being sent to that Class D address.

   When establishing a new multicast VC it is possible that one or more
   L_MULTI_RQ or L_MULTI_ADD may fail. 2. The UNI 3.1 failure cause must
   be returned in the ERR_L_RQFAILED signal from the local signalling
   entity
   message format is:




Armitage                Expires April 24th, 1996                [Page 21]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


      Data:
       ar$hrd      16 bits  Hardware type.
       ar$pro.type 16 bits  Protocol type.
       ar$pro.snap 40 bits  Optional SNAP extension to the AAL User. If the failure cause is not 49 (Quality protocol type.
       ar$hdrrsv   24 bits  Reserved. Unused by MARS control protocol.
       ar$chksum   16 bits  Checksum across entire MARS message.
       ar$extoff   16 bits  Extensions Offset.
       ar$op       16 bits  Operation code (MARS_MULTI = 2).
       ar$shtl      8 bits  Type & length of
   Service unavailable) or 51 (user cell rate not available), the
   endpoint's source ATM number.
       ar$sstl      8 bits  Type & length of source ATM subaddress.
       ar$spln      8 bits  Length of source protocol address is dropped from the set {ATM.1, ATM.2, ...,
   ATM.n} returned by the MARS.  Otherwise, the L_MULTI_RQ or
   L_MULTI_ADD should be reissued after a random delay (s)
       ar$thtl      8 bits  Type & length of 5 to 10
   seconds.  If the request fails again, another request should be
   issued after twice the previous delay has elapsed.  This process
   should be continued until the call succeeds or the multipoint VC gets
   released.

   If the initial L_MULTI_RQ fails for ATM.1, and n is greater than 1
   (i.e. the returned set target ATM number (x)
       ar$tstl      8 bits  Type & length of target ATM subaddress (y)
       ar$tpln      8 bits  Length of target multicast group address (z)
       ar$tnum     16 bits  Number of target ATM addresses contains returned (N).
       ar$seqxy    16 bits  Boolean flag x and sequence number y.
       ar$msn      32 bits  MARS Sequence Number.
       ar$sha      qoctets  source ATM number
       ar$ssa      roctets  source ATM subaddress
       ar$spa      soctets  source protocol address
       ar$tpa      zoctets  target multicast group address
       ar$tha.1    xoctets  target ATM number 1
       ar$tsa.1    yoctets  target ATM subaddress 1
       ar$tha.2    xoctets  target ATM number 2 or more addresses)
   a new L_MULTI_RQ should be immediately issued for the next
       ar$tsa.2    yoctets  target ATM subaddress 2
                 [.......]
       ar$tha.N    xoctets  target ATM number N
       ar$tsa.N    yoctets  target ATM subaddress N

   The source protocol and ATM address fields are copied directly from
   the MARS_REQUEST that this MARS_MULTI is in response to (not the set. This procedure MARS
   itself).

   ar$seqxy is repeated until an L_MULTI_RQ
   succeeds, coded with flag x in the leading bit, and sequence number
   y coded as no L_MULTI_ADDs may be issued until an initial outgoing



Armitage                Expires March 20th, 1996                [Page 20] unsigned integer in the remaining 15 bits.

          |  1st octet    |   2nd octet   |
           7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0
          +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
          |x|                 y           |
          +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   ar$tnum is an unsigned integer indicating how many pairs of
   {ar$tha,ar$tsa} (i.e. how many group member's ATM addresses) are
   present in the message. ar$msn is an unsigned 32 bit number filled in
   by the MARS before transmitting each MARS_MULTI. Its use is described
   further in section 5.1.4.




Armitage                Expires April 24th, 1996                [Page 22]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   VC is established.

   Each


   As an example, assume we have a multicast cluster using 4 byte
   protocol addresses, 20 byte ATM address for which numbers, and 0 byte ATM subaddresses.
   For n group members in a single MARS_MULTI we require a (48 + 20n)
   byte message. If we assume the default MTU of 9180 bytes, we can
   return a maximum of 456 group member's addresses in a single
   MARS_MULTI.

5.1.3   Establishing the outgoing multipoint VC.

   Following the completion of the MARS_MULTI reply the endpoint may
   establish a new point to multipoint VC, or reuse an existing one.

   If establishing a new VC, an L_MULTI_RQ failed with cause 49 or 51
   MUST be tagged rather than deleted. An L_MULTI_ADD is issued for
   these tagged addresses using ATM.1, followed
   by an L_MULTI_ADD for every member of the random delay procedure outlined
   above. set {ATM.2, ....ATM.n}
   (assuming the set is non-null). The packet is then transmitted over
   the newly created VC MAY just as it would be considered 'up' before failed L_MULTI_ADDs have been
   successfully re-issued. An endpoint MAY implement for a concurrent
   mechanism that allows data to start flowing out unicast VC.

   After transmitting the new VC even while
   failed L_MULTI_ADDs are being re-tried. (The alternative of waiting
   for each leaf node to accept packet, the connection could lead to significant
   delays in transmitting local interface holds the first packet.)

   Each VC MUST have a configurable inactivity timer associated with it.
   If the timer expires, an L_RELEASE is issued for that VC, open
   and marks it as the
   Class D address is no longer considered to have an active path out of the local host. host for any subsequent IP
   packets being sent to that Class D address.

   When establishing a new multicast VC it is possible that one or more
   L_MULTI_RQ or L_MULTI_ADD may fail.  The timer SHOULD UNI 3.0/3.1 failure cause
   must be no less than 1 minute, and a
   default of 20 minutes returned in the ERR_L_RQFAILED signal from the local
   signalling entity to the AAL User. If the failure cause is RECOMMENDED. Choice not 49
   (Quality of specific timer
   periods Service unavailable) or 51 (user cell rate not
   available), the endpoint's ATM address is beyond dropped from the scope of this document.

   VC consumption may also be reduced by endpoints noting when a new
   group's set of
   {ATM.1, ....ATM.n} matches that of a pre-existing VC
   out to another group. With careful local management, and assuming ATM.2, ..., ATM.n} returned by the
   QoS of MARS.  Otherwise, the existing VC is sufficient for both groups,
   L_MULTI_RQ or L_MULTI_ADD should be reissued after a new pt random delay of
   5 to mpt
   VC may not 10 seconds.  If the request fails again, another request should
   be necessary.  Under certain circumstances endpoints may
   decide that it is sufficient to re-use an existing issued after twice the previous delay has elapsed.  This process
   should be continued until the call succeeds or the multipoint VC whose set of
   leaf nodes is a superset of gets
   released.

   If the new group's membership (in which case
   some endpoints will receive multicast traffic initial L_MULTI_RQ fails for a layer 3 group
   they haven't joined, ATM.1, and must filter them above n is greater than 1
   (i.e. the ATM interface).
   Algorithms for performing this type returned set of optimization are not discussed
   here, and are not required for conformance with this document.

5.1.4   Tracking subsequent group updates.

   Once ATM addresses contains 2 or more addresses)
   a new VC has been established, L_MULTI_RQ should be immediately issued for the transmit side of next ATM
   address in the cluster
   member's interface needs to monitor subsequent group changes - adding
   or dropping leaf nodes as appropriate. set. This procedure is achieved by watching
   for MARS_JOIN and MARS_LEAVE messages from the MARS itself. These
   messages are described in detail in section 5.2 - at this point it repeated until an L_MULTI_RQ
   succeeds, as no L_MULTI_ADDs may be issued until an initial outgoing
   VC is
   sufficient to note that they carry:

      - The established.

   Each ATM address of a node joining or leaving a group.
      - The layer 3 address of the group(s) being joined for which an L_MULTI_RQ failed with cause 49 or left.
      - A Cluster Sequence Number (CSN) from 51
   MUST be tagged rather than deleted. An L_MULTI_ADD is issued for
   these tagged addresses using the MARS.

   MARS_JOIN and MARS_LEAVE messages arrive at each cluster member
   across ClusterControlVC. MARS_JOIN or MARS_LEAVE messages that simply random delay procedure outlined
   above.

   The VC MAY be considered 'up' before failed L_MULTI_ADDs have been
   successfully re-issued. An endpoint MAY implement a concurrent



Armitage                Expires March 20th, April 24th, 1996                [Page 21] 23]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   confirm information already held by the cluster member are used to
   track the Cluster Sequence Number, but are otherwise ignored.

5.1.4.1   Updating the active VCs.

   If a MARS_JOIN is seen


   mechanism that refers allows data to (or encompasses) a group for
   which the transmit side already has a VC open, start flowing out the new member's ATM
   address is extracted and an L_MULTI_ADD issued locally. This ensures
   that endpoints already sending VC even while
   failed L_MULTI_ADDs are being re-tried. (The alternative of waiting
   for each leaf node to a given group will immediately add accept the new member to their list of recipients.

   If a MARS_LEAVE is seen that refers connection could lead to (or encompasses) a group for
   which significant
   delays in transmitting the transmit side already has a first packet.)

   Each VC open, MUST have a configurable inactivity timer associated with it.
   If the old member's ATM
   address is extracted and timer expires, an L_MULTI_DROP L_RELEASE is issued locally. This ensures for that endpoints already sending to a given group will immediately drop
   the old member from their list of recipients. When the last leaf of a
   VC is dropped, the VC is closed completely VC, and the affected group
   Class D address is no longer has a considered to have an active path out of
   the local endpoint (the next outbound packet
   to that group's address will trigger the creation of a new VC, as
   described in sections 5.1.1 to 5.1.3).

   In an IPv4 environment any endpoint leaving 224.0.0.1 is assumed to host. The timer SHOULD be ceasing support for IP multicast operation. If no less than 1 minute, and a MARS_LEAVE
   default of 20 minutes is
   seen RECOMMENDED. Choice of specific timer
   periods is beyond the scope of this document.

   VC consumption may also be reduced by endpoints noting when a new
   group's set of {ATM.1, ....ATM.n} matches that refers of a pre-existing VC
   out to group 224.0.0.1 then another group. With careful local management, and assuming the ATM address
   QoS of the
   endpoint specified in the message MUST be removed from every
   multipoint existing VC on which it is listed as sufficient for both groups, a leaf node.

   The transmit side of the interface MUST NOT shut down an active new pt to mpt
   VC may not be necessary.  Under certain circumstances endpoints may
   decide that it is sufficient to re-use an existing VC whose set of
   leaf nodes is a group for which superset of the new group's membership (in which case
   some endpoints will receive side has just executed multicast traffic for a
   LeaveLocalGroup.  (This behaviour is consistent with layer 3 group
   they haven't joined, and must filter them above the model of
   hosts transmitting to groups regardless ATM interface).
   Algorithms for performing this type of their own membership
   status.)

   If a MARS_JOIN or MARS_LEAVE arrives with ar$pnum == 0 it carries no
   <min,max> pairs, optimization are not discussed
   here, and is only used are not required for tracking the CSN.

5.1.4.2 conformance with this document.

5.1.4   Tracking subsequent group updates.

   Once a new VC has been established, the Cluster Sequence Number.

   It is important that endpoints do not miss transmit side of the cluster
   member's interface needs to monitor subsequent group membership updates
   issued changes - adding
   or dropping leaf nodes as appropriate. This is achieved by watching
   for MARS_JOIN and MARS_LEAVE messages from the MARS over ClusterControlVC. However, itself. These
   messages are described in detail in section 5.2 - at this will happen
   from time point it is
   sufficient to time. note that they carry:

      - The ATM address of a node joining or leaving a group.
      - The layer 3 address of the group(s) being joined or left.
      - A Cluster Sequence Number is carried as an
   unsigned 32 bit value in (CSN) from the ar$msn field of many MARS messages
   (except for MARS_REQUEST MARS.

   MARS_JOIN and MARS_NAK).  It increments once for every
   transmission MARS_LEAVE messages arrive at each cluster member
   across ClusterControlVC. MARS_JOIN or MARS_LEAVE messages that simply
   confirm information already held by the MARS makes on ClusterControlVC, regardless of
   whether cluster member are used to
   track the transmission represents Cluster Sequence Number, but are otherwise ignored.

5.1.4.1   Updating the active VCs.

   If a change in MARS_JOIN is seen that refers to (or encompasses) a group for
   which the MARS database or
   not. By tracking this counter, cluster members can determine whether
   they have missed transmit side already has a previous message on ClusterControlVC, VC open, the new member's ATM
   address is extracted and possibly
   a membership change. an L_MULTI_ADD issued locally. This is then used ensures
   that endpoints already sending to trigger revalidation a given group will immediately add



Armitage                Expires March 20th, April 24th, 1996                [Page 22] 24]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   (described in section 5.1.5).

   The current CSN is copied into


   the ar$msn field of MARS messages
   being sent to cluster members, whether out ClusterControlVC or on an
   point to point VC.

   Calculations on the sequence numbers MUST be performed as unsigned 32
   bit arithmetic.

   Every cluster new member keeps its own 32 bit Host Sequence Number (HSN) to track the MARS's sequence number. Whenever their list of recipients.

   If a message is received
   that carries an ar$msn field the following processing is performed:

         Seq.diff = ar$msn - HSN

         ar$msn -> HSN
         {...process MARS message as appropriate...}

         if ((Seq.diff != 1) && (Seq.diff != 0))
            then {...revalidate group membership information...}

   The basic result MARS_LEAVE is seen that the cluster member attempts refers to keep locked
   in step with membership changes noted by the MARS. If it ever detects
   that (or encompasses) a membership change occurred (in any group) without it noticing,
   it re-validates group for
   which the membership of all groups it currently transmit side already has
   multicast VCs open to.

   The ar$msn value in an individual MARS_MULTI a VC open, the old member's ATM
   address is not used extracted and an L_MULTI_DROP issued locally. This ensures
   that endpoints already sending to update a given group will immediately drop
   the HSN until all parts old member from their list of recipients. When the MARS_MULTI (if more than 1) have
   arrived.  However, the ar$msn field in consecutive messages last leaf of a
   multi-part MARS_MULTI MUST be constant. If the ar$msn field changes
   before
   VC is dropped, the MARS_MULTI VC is closed completely received, then the entire
   MARS_MULTI MUST be discarded at and the completion affected group no
   longer has a path out of the response, and
   the MARS_REQUEST re-issued.

   The MARS is free local endpoint (the next outbound packet
   to choose an initial value that group's address will trigger the creation of CSN. When a new
   cluster member starts up it should initialise HSN VC, as
   described in sections 5.1.1 to zero. When the
   cluster member sends the MARS_JOIN 5.1.3).

   In an IPv4 environment any endpoint leaving 224.0.0.1 is assumed to register (described later), the
   HSN will
   be correctly updated ceasing support for IP multicast operation. If a MARS_LEAVE is
   seen that refers to group 224.0.0.1 then the current CSN value when ATM address of the
   endpoint receives specified in the copy of its MARS_JOIN back message MUST be removed from the MARS.

5.1.5   Revalidating every
   multipoint VC on which it is listed as a VC's leaf nodes.

   Certain events may inform a cluster member that it has incorrect
   information about the sets node.

   The transmit side of leaf nodes it should be sending to.  If the interface MUST NOT shut down an error occurs on a active VC associated with to
   a particular group, the
   cluster member initiates revalidation procedures group for that specific
   group. If which the receive side has just executed a jump
   LeaveLocalGroup.  (This behaviour is detected in consistent with the Cluster Sequence Number, this



Armitage                Expires March 20th, 1996                [Page 23]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


   initiates revalidation model of all groups to which the cluster member
   currently has open point
   hosts transmitting to multipoint VCs.

   Each open and active multipoint VC has groups regardless of their own membership
   status.)

   If a flag associated MARS_JOIN or MARS_LEAVE arrives with ar$pnum == 0 it
   called 'VC_revalidate'. This flag is checked everytime a packet carries no
   <min,max> pairs, and is
   queued only used for transmission on that VC. If tracking the flag is false, CSN.

5.1.4.2   Tracking the packet
   is transmitted and no further action Cluster Sequence Number.

   It is required.

   However, if important that endpoints do not miss group membership updates
   issued by the VC_revalidate flag MARS over ClusterControlVC. However, this will happen
   from time to time. The Cluster Sequence Number is true then carried as an
   unsigned 32 bit value in the packet is
   transmitted and a new sequence ar$msn field of events is started locally.

   Revalidation begins with re-issuing a many MARS messages
   (except for MARS_REQUEST and MARS_NAK).  It increments once for every
   transmission the group
   being revalidated.  The returned set of members {NewATM.1, NewATM.2,
   .... NewATM.n} is compared with the set already held locally.
   L_MULTI_DROPs are issued MARS makes on the group's VC for each node that appears
   in the original set of members but not in the revalidated set ClusterControlVC, regardless of
   members. L_MULTI_ADDs are issued on
   whether the group's VC for each node that
   appears transmission represents a change in the revalidated set of MARS database or
   not. By tracking this counter, cluster members but not in the original set
   of members. The VC_revalidate flag can determine whether
   they have missed a previous message on ClusterControlVC, and possibly
   a membership change. This is reset when revalidation
   concludes for the given group. Implementation specific mechanisms
   will be needed then used to flag the 'revalidation trigger revalidation
   (described in progress' state. section 5.1.5).

   The key difference between constructing a VC (section 5.1.3) and
   revalidating a VC current CSN is that packet transmission continues on copied into the open
   VC while it is ar$msn field of MARS messages
   being revalidated. This minimises the disruption sent to
   existing traffic.

   The algorithm for initiating revalidation is:

      - When a packet arrives for transmission cluster members, whether out ClusterControlVC or on a given group,
        the groups membership is revalidated if VC_revalidate == TRUE.
        Revalidation resets VC_revalidate.
      - When an event occurs that demands revalidation, every
        group has its VC_revalidate flag set TRUE at a random time
        between 1 and 10 seconds.

   Benefit: Revalidation of active groups occurs quickly, and
   essentially idle groups are revalidated as needed. Randomly
   distributed setting of VC_revalidate flag improves chances of
   staggered revalidation requests from senders when a sequence number
   jump is detected.

5.1.5.1   When leaf node drops itself.

   During the life of a multipoint VC an ERR_L_DROP may be received
   indicating that a leaf node has terminated its participation at the
   ATM level. The ATM endpoint associated with
   point to point VC.

   Calculations on the ERR_L_DROP sequence numbers MUST be
   removed from the locally held set {ATM.1, ATM.2, .... ATM.n} performed as unsigned 32
   bit arithmetic.




Armitage                Expires March 20th, April 24th, 1996                [Page 24] 25]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   associated with


   Every cluster member keeps its own 32 bit Host Sequence Number (HSN)
   to track the VC.

   After MARS's sequence number. Whenever a random period of time between 1 and 10 seconds the
   VC_revalidate flag associated with message is received
   that VC MUST be set true.

   If carries an ERR_L_RELEASE ar$msn field the following processing is received performed:

         Seq.diff = ar$msn - HSN

         ar$msn -> HSN
         {...process MARS message as appropriate...}

         if ((Seq.diff != 1) && (Seq.diff != 0))
            then the entire set {ATM.1, ATM.2,
   .... ATM.n} {...revalidate group membership information...}

   The basic result is cleared and that the VC is considered to be completely shut
   down. Further packet transmission cluster member attempts to the group served by this VC will
   result keep locked
   in step with membership changes noted by the MARS. If it ever detects
   that a new VC being established membership change occurred (in any group) without it noticing,
   it re-validates the membership of all groups it currently has
   multicast VCs open to.

   The ar$msn value in an individual MARS_MULTI is not used to update
   the HSN until all parts of the MARS_MULTI (if more than 1) have
   arrived. (If the ar$msn changes the MARS_MULTI is discarded, as
   described in section 5.1.3.

5.1.5.2   When a jump 5.1.1.)

   The MARS is detected in the free to choose an initial value of CSN.

   Section 5.1.4.2 describes how a CSN jump is detected. If When a new
   cluster member starts up it should initialise HSN to zero. When the
   cluster member sends the MARS_JOIN to register (described later), the
   HSN will be correctly updated to the current CSN jump
   is detected upon receipt value when the
   endpoint receives the copy of a MARS_JOIN or a MARS_LEAVE then every
   outgoing multicast VC MUST have its VC_revalidate flag set true at
   some random interval between 1 and 10 seconds MARS_JOIN back from when the CSN jump
   was detected.

   The only exception to this rule is if MARS.

5.1.5   Revalidating a sequence number jump is
   detected during VC's leaf nodes.

   Certain events may inform a cluster member that it has incorrect
   information about the establishment sets of leaf nodes it should be sending to.  If
   an error occurs on a new group's VC (i.e. associated with a
   MARS_MULTI reply was correctly received, but its ar$msn indicated particular group, the
   cluster member initiates revalidation procedures for that some previous MARS traffic had been missed on ClusterControlVC).
   In specific
   group. If a jump is detected in the Cluster Sequence Number, this case every
   initiates revalidation of all groups to which the cluster member
   currently has open VC, EXCEPT point to multipoint VCs.

   Each open and active multipoint VC has a flag associated with it
   called 'VC_revalidate'. This flag is checked everytime a packet is
   queued for transmission on that VC. If the flag is false, the packet
   is transmitted and no further action is required.

   However, if the one just established, MUST
   have its VC_revalidate flag set is true at some random interval between
   1 then the packet is
   transmitted and 10 seconds from when a new sequence of events is started locally.

   Revalidation begins with re-issuing a MARS_REQUEST for the CSN jump was detected.  (The VC group



Armitage                Expires April 24th, 1996                [Page 26]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   being
   established at the time revalidated.  The returned set of members {NewATM.1, NewATM.2,
   .... NewATM.n} is considered compared with the set already validated.)


5.2.   Receive side behaviour.

   A cluster member is a 'group member' (in held locally.
   L_MULTI_DROPs are issued on the sense group's VC for each node that it receives
   packets directed at a given multicast group) when its ATM address appears
   in the MARS's table entry for original set of members but not in the revalidated set of
   members. L_MULTI_ADDs are issued on the group's multicast address.
   A key function within VC for each cluster is node that
   appears in the distribution revalidated set of group
   membership information from members but not in the MARS to cluster members.

   An endpoint may wish original set
   of members. The VC_revalidate flag is reset when revalidation
   concludes for the given group. Implementation specific mechanisms
   will be needed to 'join a group' flag the 'revalidation in response progress' state.

   The key difference between constructing a VC (section 5.1.3) and
   revalidating a VC is that packet transmission continues on the open
   VC while it is being revalidated. This minimises the disruption to
   existing traffic.

   The algorithm for initiating revalidation is:

      - When a local, higher
   level request packet arrives for membership of transmission on a given group, or because the endpoint
   supports a layer 3 multicast forwarding engine that requires
        the
   ability to 'see' intra-cluster traffic in order to forward it.

   Two messages support these requirements groups membership is revalidated if VC_revalidate == TRUE.
        Revalidation resets VC_revalidate.
      - MARS_JOIN When an event occurs that demands revalidation, every
        group has its VC_revalidate flag set TRUE at a random time
        between 1 and MARS_LEAVE.
   These 10 seconds.

   Benefit: Revalidation of active groups occurs quickly, and
   essentially idle groups are sent to the MARS by endpoints revalidated as needed. Randomly
   distributed setting of VC_revalidate flag improves chances of
   staggered revalidation requests from senders when the local layer 3/ATM
   interface a sequence number
   jump is requested to join or leave detected.

5.1.5.1   When leaf node drops itself.

   During the life of a multicast group. multipoint VC an ERR_L_DROP may be received
   indicating that a leaf node has terminated its participation at the
   ATM level. The MARS
   propagates these messages back out over ClusterControlVC, to ensure ATM endpoint associated with the knowledge of ERR_L_DROP MUST be
   removed from the group's membership change locally held set {ATM.1, ATM.2, .... ATM.n}
   associated with the VC.

   After a random period of time between 1 and 10 seconds the
   VC_revalidate flag associated with that VC MUST be set true.

   If an ERR_L_RELEASE is distributed received then the entire set {ATM.1, ATM.2,
   .... ATM.n} is cleared and the VC is considered to be completely shut
   down. Further packet transmission to the group served by this VC will
   result in a
   timely fashion to other cluster members. new VC being established as described in section 5.1.3.






Armitage                Expires March 20th, April 24th, 1996                [Page 25] 27]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   Certain models of layer 3 endpoints (e.g. IP multicast routers)
   expect to be able to receive packet traffic 'promiscuously' across
   all groups.  This functionality may be emulated by allowing routers
   to request that the MARS returns them as 'wild card' members of all
   Class D addresses.  However,


5.1.5.2   When a problem inherent jump is detected in the current ATM
   model CSN.

   Section 5.1.4.2 describes how a CSN jump is that detected. If a CSN jump
   is detected upon receipt of a completely promiscuous router may exhaust the local
   reassembly resources in its ATM interface. MARS_JOIN supports or a
   generalisation to MARS_LEAVE then every
   outgoing multicast VC MUST have its VC_revalidate flag set true at
   some random interval between 1 and 10 seconds from when the notion of 'wild card' entries, enabling routers
   to limit themselves CSN jump
   was detected.

   The only exception to 'blocks' of the Class D address space. Use of this facility rule is described in greater detail in Section 8.

   A block can be as small as 1 (a single group) or as large as the
   entire multicast address space (e.g. default IPv4 'promiscuous'
   behaviour).  A block if a sequence number jump is defined as all addresses between, and
   inclusive of,
   detected during the establishment of a <min,max> address pair. A MARS_JOIN or MARS_LEAVE may
   carry multiple <min,max> pairs.

   Cluster members MUST provide ONLY new group's VC (i.e. a single <min,max> pair in each
   JOIN/LEAVE message they issue. However, they
   MARS_MULTI reply was correctly received, but its ar$msn indicated
   that some previous MARS traffic had been missed on ClusterControlVC).
   In this case every open VC, EXCEPT the one just established, MUST be able to process
   multiple <min,max> pairs in JOIN/LEAVE messages
   have its VC_revalidate flag set true at some random interval between
   1 and 10 seconds from when performing the CSN jump was detected.  (The VC
   management as described in section 5.1.4 (the interpretation being
   that
   established at the join/leave operation applies to all addresses in range from
   <min> to <max> inclusive, for every <min,max> pair).

   In RFC1112 environments a MARS_JOIN for a single group time is triggered
   by a JoinLocalGroup signal from the IP layer. considered already validated.)


5.2.   Receive side behaviour.

   A MARS_LEAVE for a
   single group cluster member is triggered by a LeaveLocalGroup signal from 'group member' (in the IP
   layer.

   Cluster members with special requirements (e.g. multicast routers)
   may issue MARS_JOINs and MARS_LEAVEs specifying sense that it receives
   packets directed at a block of given multicast group) when its ATM address
   appears in the MARS's table entry for the group's multicast address.
   A key function within each cluster is the distribution of group addresses.
   membership information from the MARS to cluster members.

   An endpoint MUST register with may wish to 'join a MARS group' in order response to become a member local, higher
   level request for membership of a cluster and be added as a leaf to ClusterControlVC.  Registration
   is covered in section 5.2.3.

   Finally, group, or because the endpoint MUST be capable of terminating unidirectional
   VCs (i.e. act as a leaf node of
   supports a UNI 3.1 point to multipoint VC,
   with zero bandwidth assigned on the return path). RFC 1755 describes layer 3 multicast forwarding engine that requires the signalling information required
   ability to terminate VCs carrying
   LLC/SNAP encapsulated 'see' intra-cluster traffic (discussed further in section 5.5).

5.2.1 Format of the order to forward it.

   Two messages support these requirements - MARS_JOIN and MARS_LEAVE Messages.

   The MARS_JOIN message is indicated MARS_LEAVE.
   These are sent to the MARS by an operation type value of 14
   (decimal). MARS_LEAVE has endpoints when the same format and operation type value of
   15 (decimal). The message format is:



Armitage                Expires March 20th, 1996                [Page 26]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


      Data:
       ar$hrd     16 bits  Hardware type (0x13)
       ar$pro     16 bits  Protocol type
       ar$shtl     8 bits  Type & length of source ATM number (q)
       ar$sstl     8 bits  Type & length of source ATM subaddress (r)
       ar$op      16 bits  Operation code (MARS_JOIN local layer 3/ATM
   interface is requested to join or MARS_LEAVE)
       ar$extoff  32 bits  Extensions Offset.
       ar$spln     8 bits  Length leave a multicast group. The MARS
   propagates these messages back out over ClusterControlVC, to ensure
   the knowledge of source protocol address (s)
       ar$tpln     8 bits  Length the group's membership change is distributed in a
   timely fashion to other cluster members.

   Certain models of layer 3 endpoints (e.g. IP multicast group address (z)
       ar$pnum    16 bits  Number of multicast group address pairs (N)
       ar$flags   16 bits  layer3grp, copy, and register flags.
       ar$cmi     16 bits  Cluster Member ID
       ar$msn     32 bits  MARS Sequence Number.
       ar$sha     qoctets  source ATM number (E.164 or ATM Forum NSAPA).
       ar$ssa     roctets  source ATM subaddress (ATM Forum NSAPA).
       ar$spa     soctets  source protocol address
       ar$min.1   zoctets  Minimum multicast group address - pair.1
       ar$max.1   zoctets  Maximum multicast group address - pair.1
                 [.......]
       ar$min.N   zoctets  Minimum multicast group address - pair.N
       ar$max.N   zoctets  Maximum multicast group address - pair.N

   ar$spln indicates routers)
   expect to be able to receive packet traffic 'promiscuously' across
   all groups.  This functionality may be emulated by allowing routers
   to request that the number MARS returns them as 'wild card' members of bytes all
   Class D addresses.  However, a problem inherent in the source endpoint's
   protocol address, and current ATM
   model is interpreted that a completely promiscuous router may exhaust the local
   reassembly resources in its ATM interface. MARS_JOIN supports a
   generalisation to the context notion of 'wild card' entries, enabling routers
   to limit themselves to 'blocks' of the protocol
   indicated by Class D address space. Use of
   this facility is described in greater detail in Section 8.



Armitage                Expires April 24th, 1996                [Page 28]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   A block can be as small as 1 (a single group) or as large as the ar$pro field.
   entire multicast address space (e.g. in default IPv4 environments ar$pro will
   be 0x800, ar$spln 'promiscuous'
   behaviour).  A block is 4, defined as all addresses between, and ar$tpln is 4.)

   The ar$flags field contains three flags:

      Bit 15  - ar$flags.layer3grp.
      Bit 14  - ar$flags.copy.
      Bit 13  - ar$flags.register.
      Bit 12  - ar$flags.punched.
      Bit 0-7 - ar$flags.sequence.

   Bits 8 to 11 are reserved and MUST be zero.

   ar$flags.sequence is set by cluster members, and MUST always be
   passed on unmodified by the MARS when retransmitting
   inclusive of, a <min,max> address pair. A MARS_JOIN or MARS_LEAVE messages. It is source specific, and may
   carry multiple <min,max> pairs.

   Cluster members MUST be ignored by
   other cluster members. Its use is described provide ONLY a single <min,max> pair in section 5.2.2.

   ar$flags.punched each
   JOIN/LEAVE message they issue. However, they MUST be zero when the MARS_JOIN or MARS_LEAVE is
   transmitted able to the MARS. Its use is process
   multiple <min,max> pairs in JOIN/LEAVE messages when performing VC
   management as described in section 5.2.2 and
   section 6.2.4.

   ar$flags.copy MUST be set to 0 when the message is 5.1.4 (the interpretation being sent
   that the join/leave operation applies to all addresses in range from a
   MARS client, and MUST be set
   <min> to 1 when the message <max> inclusive, for every <min,max> pair).

   In RFC1112 environments a MARS_JOIN for a single group is being sent triggered
   by a JoinLocalGroup signal from



Armitage                Expires March 20th, 1996                [Page 27]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995 the IP layer. A MARS_LEAVE for a MARS. (This flag
   single group is intended to support integrating triggered by a LeaveLocalGroup signal from the MARS
   function IP
   layer.

   Cluster members with one special requirements (e.g. multicast routers)
   may issue MARS_JOINs and MARS_LEAVEs specifying a block of the multicast
   group addresses.

   An endpoint MUST register with a MARS clients in your cluster. The
   destination order to become a member of an incoming MARS_JOIN can
   a cluster and be determined from its
   value.)

   ar$flags.layer3grp allows the MARS to provide the group membership
   information described further added as a leaf to ClusterControlVC.  Registration
   is covered in section 5.3. The rules for its use
   are:

      ar$flags.layer3grp 5.2.3.

   Finally, the endpoint MUST be set when the cluster member is issuing
      the MARS_JOIN a the result capable of a layer 3 multicast group being
      explicitly joined. (e.g. terminating unidirectional
   VCs (i.e. act as a result leaf node of a JoinHostGroup operation
      in an RFC1112 compliant host).

      ar$flags.layer3grp MUST be reset in each MARS_JOIN if the
      MARS_JOIN is simply the local ip/atm interface registering UNI 3.0/3.1 point to
      receive traffic multipoint VC,
   with zero bandwidth assigned on that group for its own reasons.

      ar$flags.layer3grp is ignored and MUST be treated as reset by the
      MARS for any MARS_JOIN that specifies a block covering more than a
      single group (e.g. a block join from a router ensuring their
      forwarding engines 'see' all traffic).

   ar$flags.register indicates whether return path). RFC 1755 describes
   the MARS_JOIN or MARS_LEAVE is
   being used signalling information required to register or deregister a cluster member (described terminate VCs carrying
   LLC/SNAP encapsulated traffic (discussed further in section 5.2.3). When used to join or leave specific groups the
   ar$register flag MUST be zero.

   ar$pnum indicates how many <min,max> pairs are included in the
   message. This field MUST be 1 when 5.5).

5.2.1 Format of the message is sent from a cluster
   member. A MARS MAY return a MARS_JOIN or and MARS_LEAVE with any ar$pnum
   value, including zero.  This will be explained futher in section
   6.2.4. Messages.

   The ar$cmi field MUST be zeroed by cluster members, and MARS_JOIN message is used indicated by an operation type value of 4.
   MARS_LEAVE has the MARS during cluster member registration, described in section
   5.2.3.

   ar$msn MUST be zero when transmitted by an endpoint. It is set to the
   current same format and operation type value of the Cluster Sequence Number 5. The
   message format is:

      Data:
       ar$hrd      16 bits  Hardware type.
       ar$pro.type 16 bits  Protocol type.
       ar$pro.snap 40 bits  Optional SNAP extension to protocol type.
       ar$hdrrsv   24 bits  Reserved. Unused by the MARS when the
   MARS_JOIN control protocol.
       ar$chksum   16 bits  Checksum across entire MARS message.
       ar$extoff   16 bits  Extensions Offset.
       ar$op       16 bits  Operation code (MARS_JOIN or MARS_LEAVE is retransmitted. Its use has been described
   in section 5.1.4.

   To simplify construction and parsing MARS_LEAVE).
       ar$shtl      8 bits  Type & length of MARS_JOIN and MARS_LEAVE
   messages, the following restrictions are imposed on the <min,max>
   pairs:

      Assume max(N) is the <max> field from the Nth <min,max> pair. source ATM number.
       ar$sstl      8 bits  Type & length of source ATM subaddress.



Armitage                Expires March 20th, April 24th, 1996                [Page 28] 29]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


      Assume min(N) is the <min> field from the Nth <min,max> pair.
      Assume a join/leave message arrives with K <min,max> pairs.
      The following must hold:
         max(N) < min(N+1) for 1 <= N < K
         max(N) >= min(N) for 1 <= N <= K

   In plain language, the set must specify an ascending sequence


       ar$spln      8 bits  Length of source protocol address blocks. The definition (s)
       ar$tpln      8 bits  Length of "greater" or "less than" may be
   protocol specific. In IPv4 environments the addresses are treated as
   32 bit, unsigned binary values (most significant byte first).

5.2.1.1 Important IPv4 default values.

   The JoinLocalGroup and LeaveLocalGroup operations are only valid for
   a single group. For any arbitrary multicast group address X the associated
   MARS_JOIN or MARS_LEAVE MUST specify a single pair <X, X>.
   ar$flags.layer3grp MUST be set under these circumstances.

   A router choosing to behave strictly in accordance with RFC1112 MUST
   specify the entire Class D space. The associated MARS_JOIN or
   MARS_LEAVE MUST specify a single pair <224.0.0.0, 239.255.255.255>.
   Whenever a router issues a MARS_JOIN only in order to forward IP
   traffic it MUST reset ar$flags.layer3grp.

   The use (z)
       ar$pnum     16 bits  Number of alternative <min, max> values by multicast routers is
   discussed in Section 8.

5.2.2   Retransmission of MARS_JOIN group address pairs (N)
       ar$flags    16 bits  layer3grp, copy, and MARS_LEAVE messages.

   Transient problems may result in register flags.
       ar$cmi      16 bits  Cluster Member ID
       ar$msn      32 bits  MARS Sequence Number.
       ar$sha      qoctets  source ATM number.
       ar$ssa      roctets  source ATM subaddress.
       ar$spa      soctets  source protocol address
       ar$min.1    zoctets  Minimum multicast group address - pair.1
       ar$max.1    zoctets  Maximum multicast group address - pair.1
                 [.......]
       ar$min.N    zoctets  Minimum multicast group address - pair.N
       ar$max.N    zoctets  Maximum multicast group address - pair.N

   ar$spln indicates the loss number of messages between bytes in the
   MARS source endpoint's
   protocol address, and cluster members

   A simple algorithm is used to solve this problem. Cluster members
   retransmit each MARS_JOIN and MARS_LEAVE message at regular intervals
   until they receive a copy back again, either on ClusterControlVC or interpreted in the VC on which they are sending context of the message.  At this point protocol
   indicated by the
   local endpoint can ar$pro field. (e.g. in IPv4 environments ar$pro will
   be certain that the MARS received 0x800, ar$spln is 4, and processed
   it. ar$tpln is 4.)

   The interval should be no shorter than 5 seconds, and a default value
   of 10 seconds is recommended. After 5 retransmissions the attempt
   should ar$flags field contains three flags:

      Bit 15  - ar$flags.layer3grp.
      Bit 14  - ar$flags.copy.
      Bit 13  - ar$flags.register.
      Bit 12  - ar$flags.punched.
      Bit 0-7 - ar$flags.sequence.

   Bits 8 to 11 are reserved and MUST be flagged locally as a failure. This zero.

   ar$flags.sequence is set by cluster members, and MUST always be considered as a
   passed on unmodified by the MARS failure, when retransmitting MARS_JOIN or
   MARS_LEAVE messages. It is source specific, and triggers MUST be ignored by
   other cluster members. Its use is described in section 5.2.2.

   ar$flags.punched MUST be zero when the MARS reconnection MARS_JOIN or MARS_LEAVE is
   transmitted to the MARS. Its use is described in section
   5.4.

   A 'copy' 5.2.2 and
   section 6.2.4.

   ar$flags.copy MUST be set to 0 when the message is defined as being sent from a received
   MARS client, and MUST be set to 1 when the message is being sent from
   a MARS. (This flag is intended to support integrating the MARS
   function with one of the following fields
   matching a previously transmitted MARS_JOIN/LEAVE: MARS clients in your cluster. The
   destination of an incoming MARS_JOIN can be determined from its
   value.)

   ar$flags.layer3grp allows the MARS to provide the group membership
   information described further in section 5.3. The rules for its use



Armitage                Expires March 20th, April 24th, 1996                [Page 29] 30]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


      - ar$op
      - ar$flags.register
      - ar$flags.sequence
      - ar$pnum
      - Source ATM address
      - First <min,max> pair

   In addition, a valid copy


   are:

      ar$flags.layer3grp MUST have be set when the following field values:

      - ar$flags.punched = 0
      - ar$flags.copy = 1

   The ar$flags.sequence field cluster member is never modified or checked by a MARS.
   Implementors MAY choose to utilize locally significant sequence
   number schemes, which MAY differ from one cluster member to issuing
      the next.
   In MARS_JOIN a the absence result of such schemes the default value for
   ar$flags.sequence MUST be zero.

   Careful implementations MAY have more than one outstanding
   (unacknowledged) MARS_JOIN/LEAVE at a time.

5.2.3   Cluster member registration and deregistration.

   To become a cluster member an endpoint must register with the MARS.
   This achieves two things - the endpoint is added layer 3 multicast group being
      explicitly joined. (e.g. as a leaf node result of
   ClusterControlVC, and the endpoint is assigned a 16 bit Cluster
   Member Identifier (CMI). The CMI uniquely identifies JoinHostGroup operation
      in an RFC1112 compliant host).

      ar$flags.layer3grp MUST be reset in each endpoint
   that is attached to the cluster.

   Registration with MARS_JOIN if the MARS occurs when an endpoint issues a
      MARS_JOIN
   with is simply the ar$flags.register flag set local ip/atm interface registering to one (bit 13 of the ar$flags
   field).

   The cluster member MUST include
      receive traffic on that group for its source ATM address, own reasons.

      ar$flags.layer3grp is ignored and MAY
   choose to specify MUST be treated as reset by the
      MARS for any MARS_JOIN that specifies a null source protocol address when registering.

   No protocol specific block covering more than a
      single group addresses are included in (e.g. a block join from a router ensuring their
      forwarding engines 'see' all traffic).

   ar$flags.register indicates whether the MARS_JOIN or MARS_LEAVE is
   being used to register or deregister a registration
   MARS_JOIN.

   The cluster member retransmits this MARS_JOIN (described in accordance with
   section 5.2.2 until it confirms that the MARS has received it. 5.2.3). When used to join or leave specific groups the registration MARS_JOIN is returned it contains a non-zero
   value
   ar$register flag MUST be zero.

   ar$pnum indicates how many <min,max> pairs are included in ar$cmi. the
   message. This value field MUST be noted by the cluster member, and
   used whenever circumstances require 1 when the message is sent from a cluster member's CMI.

   An endpoint may also choose to de-register, using
   member. A MARS MAY return a MARS_JOIN or MARS_LEAVE with
   ar$flags.register set. any ar$pnum
   value, including zero.  This would result in the MARS dropping the



Armitage                Expires March 20th, 1996                [Page 30]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


   endpoint from ClusterControlVC, removing all references to the member will be explained futher in the mapping database, and freeing up its CMI.

   As for registration, a deregistration request section
   6.2.4.

   The ar$cmi field MUST include the
   correct source ATM address for the be zeroed by cluster member, but MAY choose to
   specify a null source protocol address.

   The members, and is used by
   the MARS during cluster member retransmits this MARS_LEAVE registration, described in accordance with section 5.2.2 until it confirms that the MARS has received it.


5.3   Support for Layer 3 group management.

   Whilst the intention of this specification is to
   5.2.3.

   ar$msn MUST be independent of
   layer 3 issues, zero when transmitted by an attempt endpoint. It is being made set to assist the operation
   current value of
   layer 3 multicast routing protocols that need to ascertain if any
   groups have members within a cluster.

   One example is IP, where IGMP the Cluster Sequence Number by the MARS when the
   MARS_JOIN or MARS_LEAVE is used (as retransmitted. Its use has been described
   in section 2)
   simply to determine whether any other cluster members are listening
   to a group because they have higher layer applications that want to
   receive a group's traffic.

   Routers may choose to query the MARS for this information, rather
   than multicasting IGMP queries to 224.0.0.1 5.1.4.

   To simplify construction and incurring the
   associated cost parsing of setting up a VC to all systems in MARS_JOIN and MARS_LEAVE
   messages, the cluster.

   The query following restrictions are imposed on the <min,max>
   pairs:

      Assume max(N) is issued by sending a MARS_GROUPLIST_REQUEST to the MARS.
   MARS_GROUPLIST_REQUEST <max> field from the Nth <min,max> pair.
      Assume min(N) is built the <min> field from the Nth <min,max> pair.
      Assume a MARS_JOIN, but it has join/leave message arrives with K <min,max> pairs.
      The following must hold:
         max(N) < min(N+1) for 1 <= N < K
         max(N) >= min(N) for 1 <= N <= K

   In plain language, the set must specify an
   operation code ascending sequence of 20 (ar$op = 20). A



Armitage                Expires April 24th, 1996                [Page 31]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   address blocks. The definition of "greater" or "less than" may be
   protocol specific. In IPv4 environments the addresses are treated as
   32 bit, unsigned binary values (most significant byte first).

5.2.1.1 Important IPv4 default values.

   The JoinLocalGroup and LeaveLocalGroup operations are only valid for
   a single group. For any arbitrary group address X the associated
   MARS_JOIN or MARS_LEAVE MUST specify a single <min,max> pair <X, X>.
   ar$flags.layer3grp MUST be
   provided (ar$pnum = 1), and it specifies the range of groups set under these circumstances.

   A router choosing to behave strictly in which accordance with RFC1112 MUST
   specify the querying cluster member is interested. entire Class D space. The response from the MARS is associated MARS_JOIN or
   MARS_LEAVE MUST specify a MARS_GROUPLIST_REPLY, carrying single pair <224.0.0.0, 239.255.255.255>.
   Whenever a list router issues a MARS_JOIN only in order to forward IP
   traffic it MUST reset ar$flags.layer3grp.

   The use of the alternative <min, max> values by multicast groups within the specified <min,max> block that
   have Layer 3 members.  A group routers is noted
   discussed in this list if one or more Section 8.

5.2.2   Retransmission of the MARS_JOINs that generated its mapping entry MARS_JOIN and MARS_LEAVE messages.

   Transient problems may result in the loss of messages between the
   MARS
   contained and cluster members

   A simple algorithm is used to solve this problem. Cluster members
   retransmit each MARS_JOIN and MARS_LEAVE message at regular intervals
   until they receive a set ar$flags.layer3grp flag.

   MARS_GROUPLIST_REPLYs are transmitted copy back to the querying cluster
   member again, either on ClusterControlVC or
   the VC used to send on which they are sending the MARS_GROUPLIST_REQUEST.

   MARS_GROUPLIST_REPLY message.  At this point the
   local endpoint can be certain that the MARS received and processed
   it.

   The interval should be no shorter than 5 seconds, and a default value
   of 10 seconds is derived from recommended. After 5 retransmissions the MARS_MULTI, it may have
   multiple parts if needed, attempt
   should be flagged locally as a failure. This MUST be considered as a
   MARS failure, and triggers the MARS reconnection described in section
   5.4.

   A 'copy' is defined as a received in message with the following fields
   matching a similar manner.

      Data:
       ar$hrd     16 bits  Hardware type ( 19 decimal, 0x13 hex)
       ar$pro     16 bits  Protocol type previously transmitted MARS_JOIN/LEAVE:

      - ar$op
      - ar$flags.register
      - ar$flags.sequence
      - ar$pnum
      - Source ATM address
      - First <min,max> pair




Armitage                Expires March 20th, April 24th, 1996                [Page 31] 32]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


       ar$shtl     8 bits  Type & length of source ATM number (q)
       ar$sstl     8 bits  Type & length of source ATM subaddress (r)
       ar$op      16 bits  Operation code (MARS_GROUPLIST_REPLY)
       ar$extoff  32 bits  Extensions Offset.
       ar$spln     8 bits  Length of source protocol address (s)
       ar$thtl     8 bits  Unused


   In addition, a valid copy MUST have the following field values:

      - set to zero.
       ar$tstl     8 bits  Unused ar$flags.punched = 0
      - set ar$flags.copy = 1

   The ar$flags.sequence field is never modified or checked by a MARS.
   Implementors MAY choose to zero.
       ar$tpln     8 bits  Length of target multicast group address (z)
       ar$tnum    16 bits  Number of group addresses returned (N).
       ar$seqxy   16 bits  Boolean flag x and utilize locally significant sequence
   number y.
       ar$msn     32 bits  MARS Sequence Number.
       ar$sha     qoctets  source ATM number (E.164 or ATM Forum NSAPA).
       ar$ssa     roctets  source ATM subaddress (ATM Forum NSAPA).
       ar$spa     soctets  source protocol address
       ar$mgrp.1  zoctets  Group address 1
                 [.......]
       ar$mgrp.N  zoctets  Group address N

   ar$seqxy is coded as for the MARS_MULTI - multiple
   MARS_GROUPLIST_REPLY components are transmitted and received using schemes, which MAY differ from one cluster member to the same algorithm as described in section 5.1.1 for MARS_MULTI. The
   only difference is that protocol address are being returned rather
   than ATM addresses.

   As for MARS_MULTIs, if an error occurs in next.
   In the reception absence of a multi
   part MARS_GROUPLIST_REPLY such schemes the whole thing default value for
   ar$flags.sequence MUST be discarded zero.

   Careful implementations MAY have more than one outstanding
   (unacknowledged) MARS_JOIN/LEAVE at a time.

5.2.3   Cluster member registration and deregistration.

   To become a cluster member an endpoint must register with the
   MARS_GROUPLIST_REQUEST re-issued. (This includes the ar$msn value
   being constant.)

   Note that MARS.
   This achieves two things - the ability to generate MARS_GROUPLIST_REQUEST messages, endpoint is added as a leaf node of
   ClusterControlVC, and receive MARS_GROUPLIST_REPLY messages, the endpoint is not required for
   general host interface implementations. It assigned a 16 bit Cluster
   Member Identifier (CMI). The CMI uniquely identifies each endpoint
   that is optional for interfaces
   being implemented attached to support layer 3 multicast forwarding engines.
   However, this functionality MUST be supported by the MARS.


5.4   Support for redundant/backup cluster.

   Registration with the MARS entities.

   Endpoints are assumed to have been configured occurs when an endpoint issues a MARS_JOIN
   with the ATM address of
   at least ar$flags.register flag set to one MARS. Endpoints (bit 13 of the ar$flags
   field).

   The cluster member MUST include its source ATM address, and MAY
   choose to maintain specify a table of ATM
   addresses, representing alternative MARSs that will be contacted null source protocol address when registering.

   No protocol specific group addresses are included in
   the event that normal operation a registration
   MARS_JOIN.

   The cluster member retransmits this MARS_JOIN in accordance with
   section 5.2.2 until it confirms that the original MARS is deemed to
   have failed. It is assumed that this table orders has received it.

   When the ATM addresses registration MARS_JOIN is returned it contains a non-zero
   value in descending order of preference. ar$cmi. This value MUST be noted by the cluster member, and
   used whenever circumstances require the cluster member's CMI.

   An endpoint will typically decide there are problems may also choose to de-register, using a MARS_LEAVE with
   ar$flags.register set. This would result in the MARS
   when:




Armitage                Expires March 20th, 1996                [Page 32]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


      - It fails to establish a point to point VC dropping the
   endpoint from ClusterControlVC, removing all references to the MARS.
      - MARS_REQUESTs fail (section 5.1.1).
      - MARS_JOIN/MARS_LEAVEs fail (section 5.2.2).
      - It has not received a MARS_REDIRECT_MAP member
   in the last 4 minutes.

   (If it is able mapping database, and freeing up its CMI.

   As for registration, a deregistration request MUST include the
   correct source ATM address for the cluster member, but MAY choose to discern which connection represents
   ClusterControlVC, it may also use connection failures on
   specify a null source protocol address.




Armitage                Expires April 24th, 1996                [Page 33]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   The cluster member retransmits this VC to
   indicate problems MARS_LEAVE in accordance with
   section 5.2.2 until it confirms that the MARS).

5.4.1   First response to MARS problems.

   The first response has received it.


5.3   Support for Layer 3 group management.

   Whilst the intention of this specification is to assume a transient problem with the MARS be independent of
   layer 3 issues, an attempt is being used at made to assist the time. The cluster member should wait a random
   period operation of time between 1 and 10 seconds before attempting
   layer 3 multicast routing protocols that need to re-
   connect and re-register with the MARS. If the registration MARS_JOIN ascertain if any
   groups have members within a cluster.

   One example is successful then:

      The IP, where IGMP is used (as described in section 2)
   simply to determine whether any other cluster member MUST then proceed members are listening
   to rejoin every a group that
      its local because they have higher layer protocol(s) have joined. It is recommended applications that want to
   receive a random delay between 1 group's traffic.

   Routers may choose to query the MARS for this information, rather
   than multicasting IGMP queries to 224.0.0.1 and 10 seconds be inserted before
      attempting each MARS_JOIN.

      The cluster member MUST initiate incurring the revalidation
   associated cost of every
      multicast group it was setting up a VC to all systems in the cluster.

   The query is issued by sending a MARS_GROUPLIST_REQUEST to (as though the MARS.
   MARS_GROUPLIST_REQUEST is built from a sequence number
      jump had been detected, section 5.1.5). MARS_JOIN, but it has an
   operation code of 10. The rejoin and revalidation procedure must not disrupt first <min,max> pair will be used by the cluster
      member's use of multipoint VCs that were already open at
   MARS to identify the time range of groups in which the MARS failure.

   If re-registration querying cluster
   member is interested. Any additional <min,max> pairs will be ignored.
   A request with ar$pnum = 0 will be ignored.

   The response from the current MARS fails, and there are no
   backup MARS addresses configured, is a MARS_GROUPLIST_REPLY, carrying a list
   of the cluster member MUST wait for at
   least 1 minute before repeating multicast groups within the re-registration procedure. It is
   RECOMMENDED specified <min,max> block that the cluster member signals an error condition
   have Layer 3 members.  A group is noted in
   some locally significant fashion.

   This procedure may repeat until network administrators manually
   intervene this list if one or more
   of the MARS_JOINs that generated its mapping entry in the current MARS returns to normal operation.

5.4.2   Connecting to
   contained a backup MARS.

   If set ar$flags.layer3grp flag.

   MARS_GROUPLIST_REPLYs are transmitted back to the re-registration with querying cluster
   member on the current MARS fails, and other MARS
   addresses has been configured, the next MARS address on the list is
   chosen VC used to be the current MARS, and the cluster member immediately
   restarts send the re-registration procedure described in section 5.4.1. If
   this MARS_GROUPLIST_REQUEST.

   MARS_GROUPLIST_REPLY is succesful the cluster member will resume normal operation
   using derived from the new MARS. MARS_MULTI but with ar$op =
   11. It may have multiple parts if needed, and is RECOMMENDED that the cluster member signals received in a warning
   similar manner to a MARS_MULTI.

      Data: ar$hrd      16 bits  Hardware type.
       ar$pro.type 16 bits  Protocol type.
       ar$pro.snap 40 bits  Optional SNAP extension to protocol type.
       ar$hdrrsv   24 bits  Reserved. Unused by MARS control protocol.
       ar$chksum   16 bits  Checksum across entire MARS message.
       ar$extoff   16 bits  Extensions Offset.
       ar$op       16 bits  Operation code (MARS_GROUPLIST_REPLY).
       ar$shtl      8 bits  Type & length of this condition in some locally significant fashion. source ATM number.



Armitage                Expires March 20th, April 24th, 1996                [Page 33] 34]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   If the attempt at re-registration with the new MARS fails, the
   cluster member MUST wait for at least 1 minute before chosing the
   next MARS


       ar$sstl      8 bits  Type & length of source ATM subaddress.
       ar$spln      8 bits  Length of source protocol address in the table and repeating the procedure. If the
   end (s)
       ar$thtl      8 bits  Unused - set to zero.
       ar$tstl      8 bits  Unused - set to zero.
       ar$tpln      8 bits  Length of the table has been reached, the cluster member starts again at
   the top target multicast group address (z)
       ar$tnum     16 bits  Number of the table (which should be the original group addresses returned (N).
       ar$seqxy    16 bits  Boolean flag x and sequence number y.
       ar$msn      32 bits  MARS that the
   cluster member started with).

   In Sequence Number.
       ar$sha      qoctets  source ATM number.
       ar$ssa      roctets  source ATM subaddress.
       ar$spa      soctets  source protocol address
       ar$mgrp.1   zoctets  Group address 1
                 [.......]
       ar$mgrp.N   zoctets  Group address N

   ar$seqxy is coded as for the worst case scenario this will result in cluster members
   looping through their table of possible MARS addresses until network
   administrators manually intervene.

5.4.3   Dynamic backup lists, MARS_MULTI - multiple
   MARS_GROUPLIST_REPLY components are transmitted and soft redirects.

   To support some level of autoconfiguration, a MARS message received using
   the same algorithm as described in section 5.1.1 for MARS_MULTI. The
   only difference is defined that allows protocol address are being returned rather
   than ATM addresses.

   As for MARS_MULTIs, if an error occurs in the current MARS to broadcast on ClusterControlVC a table reception of backup MARS addresses. When this message is received, cluster
   members that maintain a list of backup MARS addresses MUST insert
   this information at multi
   part MARS_GROUPLIST_REPLY the top of their locally held list (i.e. whole thing MUST be discarded and the
   information provided by
   MARS_GROUPLIST_REQUEST re-issued. (This includes the MARS has a higher preference than
   addresses ar$msn value
   being constant.)

   Note that may have been manually configured into the cluster
   member).

   The message ability to generate MARS_GROUPLIST_REQUEST messages,
   and receive MARS_GROUPLIST_REPLY messages, is MARS_REDIRECT_MAP. not required for
   general host interface implementations. It is based on a single MARS_MULTI,
   but with an operation type code of 22 decimal. The source hardware
   address information optional for interfaces
   being implemented to support layer 3 multicast forwarding engines.
   However, this functionality MUST be that of the MARS, and supported by the source protocol
   address field MUST be null (ar$spln = 0, and no space allocated).
   The target protocol MARS.


5.4   Support for redundant/backup MARS entities.

   Endpoints are assumed to have been configured with the ATM address MUST be null (ar$tpln = 0, and no space
   allocated). If a multi-part MARS_REDIRECT_MAP begins arriving it
   should be reassembled and accepted. If of
   at least one MARS. Endpoints MAY choose to maintain a part is lost, the entire
   message should simply table of ATM
   addresses, representing alternative MARSs that will be discarded.

   This message is transmitted regularly by contacted in
   the MARS (it MUST be
   transmitted at least every 2 minutes, it is RECOMMENDED event that it is
   transmitted every 1 minute).

   In addition to keeping cluster members updated normal operation with the recommended
   list of backup MARSs, the MARS_REDIRECT_MAP is used to force cluster
   members to 'soft redirect' from one original MARS is deemed to another. If
   have failed. It is assumed that this table orders the first ATM
   address contained addresses
   in a MARS_REDIRECT_MAP is not the address descending order of preference.

   An endpoint will typically decide there are problems with the MARS currently being used by a cluster member, the cluster member
   MUST initiate the following:
   when:

      - open It fails to establish a point to point VC to the first ATM address. MARS.
      - attempt a registration MARS_REQUESTs fail (section 5.2.3).

   If the registration succeeds, the cluster member shuts down its point
   to point VC to the current MARS (if it had one open), and then
   proceeds to use the newly opened point to point VC as its connection 5.1.1).
      - MARS_JOIN/MARS_LEAVEs fail (section 5.2.2).



Armitage                Expires March 20th, April 24th, 1996                [Page 34] 35]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   to


      - It has not received a MARS_REDIRECT_MAP in the 'current MARS'. The cluster member does NOT attempt to rejoin
   the groups last 4 minutes
      (section 5.4.3).

   (If it is a member of, or revalidate groups able to discern which connection represents
   ClusterControlVC, it may also use connection failures on this VC to
   indicate problems with the MARS).

5.4.1   First response to MARS problems.

   The first response is currently
   sending to.

   This is termed to assume a 'soft redirect' because it avoids transient problem with the extra
   rejoining and revalidation processing that occurs when a MARS failure
   is
   being recovered from. It assumes some external synchronisation
   mechanisms exist between used at the old time. The cluster member should wait a random
   period of time between 1 and new MARS - mechanisms that are
   outside 10 seconds before attempting to re-
   connect and re-register with the scope of this specification.

   Some level of trust MARS. If the registration MARS_JOIN
   is required before initiating a soft redirect. A successful then:

      The cluster member MUST check then proceed to rejoin every group that
      its local higher layer protocol(s) have joined. It is recommended
      that a random delay between 1 and 10 seconds be inserted before
      attempting each MARS_JOIN.

      The cluster member MUST initiate the calling party at the other end revalidation of every
      multicast group it was sending to (as though a sequence number
      jump had been detected, section 5.1.5).

      The rejoin and revalidation procedure must not disrupt the VC on which cluster
      member's use of multipoint VCs that were already open at the MARS_REDIRECT_MAP arrived (supposedly
   ClusterControlVC) is in fact time
      of the node it trusts as MARS failure.

   If re-registration with the current MARS.

   Additional applications of this function MARS fails, and there are for further study.


5.5  Data path LLC/SNAP encapsulations.

   The following LLC/SNAP encapsulation method is no
   backup MARS addresses configured, the default to be used
   when multicasting layer 3 packets.

      [0xAA-AA-03][0x00-00-5E][0x00-01][Extended Layer 3 packet]

   (The OUI of 0x00-00-5E belongs to IANA. The PID of 0x00-01 under cluster member MUST wait for at
   least 1 minute before repeating the
   IANA space indicates that a specially extended layer 3 packet is
   being carried.)

   The extended layer 3 packet re-registration procedure. It is encoded
   RECOMMENDED that the cluster member signals an error condition in
   some locally significant fashion.

   This procedure may repeat until network administrators manually
   intervene or the following manner:

      [pkt$cmi][pkt$pro][Original Layer 3 packet]

   The first 2 octets (pkt$cmi) carry current MARS returns to normal operation.

5.4.2   Connecting to a backup MARS.

   If the Cluster Member ID (CMI)
   assigned when an endpoint registers re-registration with the current MARS (section 5.2.3).
   The second 2 octets (pkt$pro) indicate fails, and other MARS
   addresses has been configured, the protocol type of next MARS address on the
   packet carried in list is
   chosen to be the remainder of current MARS, and the payload. This is copied from cluster member immediately
   restarts the ar$pro field used re-registration procedure described in section 5.4.1. If
   this is succesful the MARS control messages.

   For example, an IPv4 packet would be transmitted as:

      [0xAA-AA-03][0x00-00-5E][0x00-01][pkt$cmi][0x800][IPv4 packet]

   The CMI cluster member will resume normal operation
   using the new MARS. It is copied into every multicast packet transmitted with RECOMMENDED that the
   above LLC/SNAP header.  When an endpoint interface receives cluster member signals
   a packet warning of this condition in some locally significant fashion.

   If the attempt at re-registration with the LLC/SNAP header shown above it compares new MARS fails, the CMI field with
   its own Cluster Member ID for that protocol.  The packet is discarded
   silently if they match. Otherwise the packet is accepted for



Armitage                Expires March 20th, April 24th, 1996                [Page 35] 36]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   processing by


   cluster member MUST wait for at least 1 minute before chosing the local protocol entity identified by
   next MARS address in the pkt$pro
   field.

   (This approach is required to allow table and repeating the filtering out procedure. If the
   end of reflected
   packets that are possible when a group is being supported by an MCS.)

   The different LLC/SNAP codepoints for unicast and multicast packet
   transmission allows a single IP/ATM interface to support both by
   demuxing on the LLC/SNAP header.


6. The MARS in greater detail.

   Section 5 implies a lot about table has been reached, the MARS's basic behaviour as observed
   by cluster members. This section summarises member starts again at
   the behaviour top of the table (which should be the original MARS
   for groups that are VC mesh based, the
   cluster member started with).

   In the worst case scenario this will result in cluster members
   looping through their table of possible MARS addresses until network
   administrators manually intervene.

5.4.3   Dynamic backup lists, and describes how soft redirects.

   To support some level of autoconfiguration, a MARSs
   behaviour changes when an MCS MARS message is registered defined
   that allows the current MARS to support broadcast on ClusterControlVC a group.

   The table
   of backup MARS addresses. When this message is intended to be received, cluster
   members that maintain a multiprotocol entity - all its mapping
   tables, CMIs, and control VCs list of backup MARS addresses MUST be managed within insert
   this information at the context top of their locally held list (i.e. the
   information provided by the ar$pro field in incoming MARS messages. For example, a MARS
   supports completely separate ClusterControlVCs for each layer 3
   protocol (ar$pro type) that it is registering members for. If has a MARS
   receives messages with an ar$pro type higher preference than
   addresses that it does not support, may have been manually configured into the cluster
   member).

   The message is dropped.

   In general the MARS treats protocol addresses as arbitrary byte
   strings. For example, the MARS will not apply IPv4 specific 'class'
   checks to addresses supplied under ar$pro = 0x800. MARS_REDIRECT_MAP. It is sufficient
   for based on the MARS to simply assume that endpoints know how to interpret MARS_MULTI
   message, with the protocol addresses that they are establishing and releasing
   mappings for.

6.1 Basic interface to Cluster members.

   The following MARS messages are used or required changes:

      - ar$tpln field replaced by cluster members:

      11   MARS_REQUEST ar$redirf.
      - ar$spln field reserved.
      - ar$tpa and ar$spa eliminated.

   MARS_REDIRECT_MAP has an operation type code of 12   MARS_MULTI
      14   MARS_JOIN
      15   MARS_LEAVE decimal.

      Data:
       ar$hrd      16   MARS_NAK
      20   MARS_GROUPLIST_REQUEST
      21   MARS_GROUPLIST_REPLY
      22   MARS_REDIRECT_MAP






Armitage                Expires March 20th, 1996                [Page 36]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


6.1.1  Response bits  Hardware type.
       ar$pro.type 16 bits  Protocol type.
       ar$pro.snap 40 bits  Optional SNAP extension to MARS_REQUEST.

   Except as described in section 6.2, if a MARS_REQUEST arrives whose protocol type.
       ar$hdrrsv   24 bits  Reserved. Unused by MARS control protocol.
       ar$chksum   16 bits  Checksum across entire MARS message.
       ar$extoff   16 bits  Extensions Offset.
       ar$op       16 bits  Operation code (MARS_REDIRECT_MAP).
       ar$shtl      8 bits  Type & length of source ATM number.
       ar$sstl      8 bits  Type & length of source ATM subaddress.
       ar$spln      8 bits  Length of source protocol address does not match that (s)
       ar$thtl      8 bits  Type & length of any registered Cluster
   member the message MUST be dropped and ignored.

6.1.2  Response to MARS_JOIN target ATM number (x)
       ar$tstl      8 bits  Type & length of target ATM subaddress (y)
       ar$redirf    8 bits  Flag controlling client redirect behaviour.
       ar$tnum     16 bits  Number of MARS addresses returned (N).
       ar$seqxy    16 bits  Boolean flag x and MARS_LEAVE.

   When a registration MARS_JOIN arrives (described in section 5.2.3)
   the sequence number y.
       ar$msn      32 bits  MARS performs the following actions:

      - Adds the node to ClusterControlVC.
      - Allocates a new Cluster Member ID (CMI).
      - Inserts Sequence Number.
       ar$sha      qoctets  source ATM number



Armitage                Expires April 24th, 1996                [Page 37]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


       ar$ssa      roctets  source ATM subaddress
       ar$tha.1    xoctets  ATM number for MARS 1
       ar$tsa.1    yoctets  ATM subaddress for MARS 1
       ar$tha.2    xoctets  ATM number for MARS 2
       ar$tsa.2    yoctets  ATM subaddress for MARS 2
                 [.......]
       ar$tha.N    xoctets  ATM number for MARS N
       ar$tsa.N    yoctets  ATM subaddress for MARS N

   The source ATM address field(s) MUST identify the new CMI into originating MARS.
   A multi-part MARS_REDIRECT_MAP may be transmitted and reassembled
   using the ar$cmi ar$seqxy field of the MARS_JOIN.
      - Retransmits in the MARS_JOIN back privately. same manner as a multi-part
   MARS_MULTI (section 5.1.1). If the node is already a registered member of failure occurs during the cluster associated
   with reassembly
   of a multi-part MARS_REDIRECT_MAP (a part lost, reassembly timeout,
   or illegal MARS Sequence Number jump) the specified protocol type then its existing CMI entire message MUST be
   discarded.

   This message is simply
   copied into the MARS_JOIN, and the MARS_JOIN retransmitted back to
   the node.  A single node may register multiple times if it supports
   multiple layer 3 protocols. The CMIs allocated transmitted regularly by the MARS for each
   such registration may or may not be the same.

   The retransmitted registration MARS_JOIN must NOT (it MUST be sent on
   ClusterControlVC.  If a cluster member issues a deregistration
   MARS_LEAVE
   transmitted at least every 2 minutes, it too is retransmitted privately.

   Non-registration MARS_JOIN and MARS_LEAVE messages are ignored if
   they arrive from a node RECOMMENDED that it is not registered as a
   transmitted every 1 minute).

   The MARS_REDIRECT_MAP is also used to force cluster member.

   Except as described in section 6.2.4, non-registration MARS_JOIN and
   MARS_LEAVE messages are retransmitted on ClusterControlVC exactly as
   they arrived (although with ar$flags.copy set members to 1), after performing
   any required database updates. The shift
   from one MARS retransmits MARS_JOIN and
   MARS_LEAVE messages even if they result in no change to another. If the database.

   MARS_JOIN or MARS_LEAVE messages MUST arrive at ATM address of the first MARS with
   ar$flags.copy set
   contained in a MARS_REDIRECT_MAP table is not the address of cluster
   member's current MARS the client MUST 'redirect' to 0, otherwise the message new MARS. The
   ar$redirf field controls how the redirection occurs.

   ar$redirf has the following format:

                7 6 5 4 3 2 1 0
               +-+-+-+-+-+-+-+-+
               |x|             |
               +-+-+-+-+-+-+-+-+

               If Bit 7 (the most significant bit) of ar$redirf is silently ignored.
   All outgoing MARS_JOIN or MARS_LEAVE messages have ar$flags.copy set
   to 1.

   ar$flags.layer3grp (section 5.3) 1
               then the cluster member MUST be ignored (and treated as
   reset) for MARS_JOINs specifying more than a single group. If perform a
   MARS_JOIN is received that contains more than one <min,max> pair, 'hard' redirect.
               Having installed the new table of MARS MUST ignore addresses carried
               by the second MARS_REDIRECT_MAP, the cluster member re-registers
               with the MARS now at the top of the table using the
               mechanism described in sections 5.4.1 and subsequent pairs.

   An additional IPv4 specific behaviour exists 5.4.2.

               If Bit 7 of ar$redirf is 0 then the cluster member MUST
               perform a 'soft' redirect, beginning with the following
               actions:

      - if open a node issues point to point VC to the first ATM address.
      - attempt a
   MARS_LEAVE for address "224.0.0.1" (the 'all systems' group) it is registration (section 5.2.3).




Armitage                Expires March 20th, April 24th, 1996                [Page 37] 38]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   assumed


   If the registration succeeds, the cluster member shuts down its point
   to have ceased multicast support completely. All references point VC to this node MUST be eliminated from any other IPv4 the current MARS (if it had one open), and then
   proceeds to use the newly opened point to point VC as its connection
   to the 'current MARS'. The cluster member does NOT attempt to rejoin
   the groups it is a member of in the database. However, the endpoint of, or revalidate groups it is NOT released as currently
   sending to.

   This is termed a
   leaf node from ClusterControlVC (this only 'soft redirect' because it avoids the extra
   rejoining and revalidation processing that occurs upon receipt of when a
   deregistration MARS_LEAVE).

   If the MARS receives a deregistration MARS_LEAVE (described in
   section 5.2.3) that member's ATM address MUST be removed from all
   groups for which it may have joined, dropped from ClusterControlVC,
   and the CMI released.

   If failure
   is being recovered from. It assumes some external synchronisation
   mechanisms exist between the old and new MARS receives an ERR_L_RELEASE on ClusterControlVC indicating - mechanisms that are
   outside the scope of this specification.

   Some level of trust is required before initiating a soft redirect. A
   cluster member has disconnected, that member's ATM address MUST be removed from all groups for which it may have joined, and the
   CMI released.

6.1.3  Generating MARS_REDIRECT_MAP.

   A MARS_REDIRECT_MAP message (described in section 5.4.3) MUST be
   regularly transmitted on ClusterControlVC.  It is RECOMMENDED check that
   this occur every 1 minute, and it MUST occur the calling party at least every 2
   minutes. If the MARS has no knowledge of other backup MARSs serving
   the cluster, it MUST include its own address as end of
   the only entry in VC on which the MARS_REDIRECT_MAP message (in addition to filling arrived (supposedly
   ClusterControlVC) is in fact the source
   address fields).

   The design and use node it trusts as the current MARS.

   Additional applications of backup MARS entities this function are for further study.

5.5  Data path LLC/SNAP encapsulations.

   An extended encapsulation scheme is beyond required to support the scope filtering
   of
   this document, possible reflected packets (section 3.3).

   Two LLC/SNAP codepoints are allocated from the IANA OUI space. These
   support two different mechanisms for detecting reflected packets.
   They are called Type #1 and will Type #2 multicast encapsulations.

   Type #1

      [0xAA-AA-03][0x00-00-5E][0x00-01][Type #1 Extended Layer 3 packet]
          LLC         OUI        PID

   Type #2

      [0xAA-AA-03][0x00-00-5E][0x00-04][Type #2 Extended Layer 3 packet]
          LLC         OUI        PID

   For conformance with this document MARS clients:

      MUST transmit data using Type #1 encapsulation.

      MUST be covered in future work.

6.1.4  Cluster Sequence Numbers. able to correctly receive traffic using Type #1 OR Type #2
      encapsulation.

      MUST NOT transmit using Type #2 encapsulation.




Armitage                Expires April 24th, 1996                [Page 39]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


5.5.1 Type #1 encapsulation.

   The Type #1 Extended layer 3 packet carries within it a copy of the
   source's Cluster Sequence Number (CSN) is described in section 5.1.4, Member ID (CMI) and
   is carried in either the ar$msn field 'short form' or 'long
   form' of MARS messages being sent the protocol type as appropriate (section 4.3).

   When carrying packets belonging to cluster
   members (either out ClusterControlVC or on an individual VC).  The
   MARS increments protocols with valid short form
   representations the CSN every time a message [Type #1 Extended Layer 3 packet] is sent on
   ClusterControlVC. encoded as:

      [pkt$cmi][pkt$pro][Original Layer 3 packet]
        2octet   2octet        N octet

   The current CSN first 2 octets (pkt$cmi) carry the CMI assigned when an endpoint
   registers with the MARS (section 5.2.3). The second 2 octets
   (pkt$pro) indicate the protocol type of the packet carried in the
   remainder of the payload. This is copied into from the ar$msn ar$pro field of used
   in the MARS messages being sent control messages.

   When carrying packets belonging to cluster members, whether out
   ClusterControlVC or on protocols that only have a private VC.

   A MARS should be carefully designed to minimise the possibility of long
   form representation (pkt$pro = 0x80) the CSN jumping unecessarily. Under normal operation only cluster
   members affected by transient link problems will miss CSN updates and overhead SHALL be forced further
   extended to revalidate. If carry the MARS itself glitches, it will be
   innundated with requests 5 byte ar$pro.snap field (with padding for a period as every cluster member
   attempts to revalidate.

   Calculations on the CSN MUST be performed as unsigned 32
   bit
   arithmetic.



Armitage                Expires March 20th, 1996                [Page 38]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


   One implication of this mechanism alignment). The encoded form SHALL be:

      [pkt$cmi][0x00-80][ar$pro.snap][padding][Original Layer 3 packet]
        2octet   2octet   5 octets   3 octets        N octet


   The CMI is that copied into the MARS should serialize
   its processing pkt$cmi field of 'simultaneous' MARS_REQUEST, MARS_JOIN and
   MARS_LEAVE messages. Join and Leave operations should be queued
   within the MARS along every outgoing Type #1
   packet.  When an endpoint interface receives an AAL_SDU with MARS_REQUESTS, and not processed until all the reply packets of a preceeding MARS_REQUEST have been transmitted.
   The transmission of MARS_REDIRECT_MAP should also be similarly
   queued.

   (The regular transmission of MARS_REDIRECT_MAP serves a secondary
   purpose of allowing cluster members to track
   LLC/SNAP codepoint indicating Type #1 encapsulation it compares the CSN, even
   CMI field with its own Cluster Member ID for the indicated protocol.
   The packet is discarded silently if they
   miss an earlier MARS_JOIN or MARS_LEAVE.)

6.2   MARS interface to Multicast Servers (MCS).

   When match. Otherwise the MARS returns packet
   is accepted for processing by the actual addresses local protocol entity identified by
   the pkt$pro (and possibly SNAP) field(s).

   Where a protocol has valid short and long forms of group members, identification,
   receivers MAY choose to additionally recognise the
   endpoint behaviour described in section 5 results in all groups being
   supported by meshes long form.

5.5.2 Type #2 encapsulation.

   Future developments may enable direct multicasting of point AAL_SDUs beyond
   cluster boundaries. Expanding the set of possible sources in this way
   may cause the CMI to multipoint VCs. However, when MCSs
   register become an inadequate parameter with which to support particular
   detect reflected packets.  A larger source identification field may
   be required.

   The Type #2 Extended layer 3 multicast groups packet carries within it an 8 octet
   source ID field and either the MARS
   modifies its use 'short form' or 'long form' of various MARS messages to fool endpoints into
   using the MCS instead.



Armitage                Expires April 24th, 1996                [Page 40]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   protocol type as appropriate (section 4.3).  The following MARS messages are associated with interaction between form and content of
   the MARS source ID field is currently unspecified, and MCSs.

      13   MARS_MSERV
      17   MARS_UNSERV
      18   MARS_SJOIN
      19   MARS_SLEAVE

   The following is not relevant to
   any MARS messages are treated client built in a slightly different
   manner when MCSs have registered to support certain group addresses:

      11   MARS_REQUEST
      14   MARS_JOIN
      15   MARS_LEAVE

   A MARS must keep two sets of mappings for each conformance with this document. Received
   Type #2 encapsulated packets MUST always be accepted and passed up to
   the higher layer 3 group using
   MCS support.  The original {layer 3 address, ATM.1, ATM.2, ... ATM.n}
   mapping (now termed indicated by the 'host map', although it includes routers) protocol identifier.

   When carrying packets belonging to protocols with valid short form
   representations the [Type #2 Extended Layer 3 packet] is
   augmented by encoded as:

      [8 octet sourceID][ar$pro.type][Null pad][Original Layer 3 packet]
                           2octets     2octets

   When carrying packets belonging to protocols that only have a parallel {layer long
   form representation (pkt$pro = 0x80) the overhead SHALL be further
   extended to carry the 5 byte ar$pro.snap field (with padding for 32
   bit alignment). The encoded form SHALL be:

      [8 octet sourceID][ar$pro.type][ar$pro.snap][Null pad][Layer 3 address, server.1, server.2, ....
   server.K} mapping (the 'server map'). It is assumed
      packet]
                           2octets      5octets      1octet

   (Note that no ATM
   addresses appear in both this case the server padding after the SNAP field is 1 octet
   rather than the 3 octets used in Type #1.)

   Where a protocol has valid short and host maps for long forms of identification,
   receivers MAY choose to additionally recognise the same
   multicast group. Typically K will be 1, but it long form.

   (Future documents may specify the contents of the source ID field.
   This will only be larger if
   multiple MCSs relevant to implementations sending Type #2
   encapsulated packets, as they are configured the only entities that need to support a given group. be
   concerned about detecting reflected Type #2 packets.)

5.5.3 A Type #1 example.

   An IPv4 packet (fully identified by an Ethertype of 0x800, therefore
   requiring 'short form' protocol type encoding) would be transmitted
   as:

      [0xAA-AA-03][0x00-00-5E][0x00-01][pkt$cmi][0x800][IPv4 packet]

      The MARS also maintains different LLC/SNAP codepoints for unicast and multicast packet
      transmission allows a point single IPv4/ATM interface to multipoint VC out to any MCSs
   registered with it, called ServerControlVC (section 6.2.3). This
   serves an analogous role to ClusterControlVC, allowing support both by
      demuxing on the MARS to LLC/SNAP header.








Armitage                Expires March 20th, April 24th, 1996                [Page 39] 41]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   update


6. The MARS in greater detail.

   Section 5 implies a lot about the MCSs with group membership changes MARS's basic behaviour as they occur. A observed
   by cluster members. This section summarises the behaviour of the MARS
   MUST also send its regular MARS_REDIRECT_MAP transmissions on both
   ServerControlVC
   for groups that are VC mesh based, and ClusterControlVC.

6.2.1   Response to describes how a MARS_REQUEST if MARSs
   behaviour changes when an MCS is registered.

   When registered to support a group.

   The MARS is intended to be a multiprotocol entity - all its mapping
   tables, CMIs, and control VCs MUST be managed within the context of
   the ar$pro field in incoming MARS receives messages. For example, a MARS_REQUEST MARS
   supports completely separate ClusterControlVCs for an address each layer 3
   protocol that has both
   host and server maps it generates is registering members for. If a response based on MARS receives
   messages with an ar$pro that it does not support, the identity of message is
   dropped.

   In general the request's source. If MARS treats protocol addresses as arbitrary byte
   strings. For example, the requestor MARS will not apply IPv4 specific 'class'
   checks to addresses supplied under ar$pro = 0x800.  It is a member of the server map sufficient
   for the requested group then the MARS returns to simply assume that endpoints know how to interpret
   the contents of protocol addresses that they are establishing and releasing
   mappings for.

   The MARS requires control messages to carry the
   host map originator's identity
   in a sequence of one or more MARS_MULTIs.  Otherwise, if the source is ATM address field(s). Messages that arrive with an
   empty ATM Number field are silently discarded prior to any other
   processing by the MARS. (Only the ATM Number field needs to be
   checked. An empty ATM Number field combined with a non-empty ATM
   Subaddress field does not represent a valid ATM address.)

   (Some example pseudo-code for a MARS can be found in Appendix F.)

6.1 Basic interface to Cluster members.

   The following MARS messages are used or required by cluster member, members:

      1   MARS_REQUEST
      2   MARS_MULTI
      4   MARS_JOIN
      5   MARS_LEAVE
      6   MARS_NAK
      10   MARS_GROUPLIST_REQUEST
      11   MARS_GROUPLIST_REPLY
      12   MARS_REDIRECT_MAP

6.1.1  Response to MARS_REQUEST.

   Except as described in section 6.2, if a MARS_REQUEST arrives whose
   source ATM address does not match that of any registered Cluster



Armitage                Expires April 24th, 1996                [Page 42]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   member the message MUST be dropped and ignored.

6.1.2  Response to MARS_JOIN and MARS_LEAVE.

   When a registration MARS_JOIN arrives (described in section 5.2.3)
   the MARS returns performs the contents of following actions:

      - Adds the server map in node to ClusterControlVC.
      - Allocates a sequence new Cluster Member ID (CMI).
      - Inserts the new CMI into the ar$cmi field of one or more MARS_MULTIs. the MARS_JOIN.
      - Retransmits the MARS_JOIN back privately.

   If the
   source node is neither a cluster member, nor already a registered member of the server map
   for the group, cluster associated
   with the request specified protocol type then its existing CMI is dropped and ignored.

   Servers use simply
   copied into the host map to establish a basic distribution VC for MARS_JOIN, and the
   group. Cluster members will establish outgoing multipoint VCs MARS_JOIN retransmitted back to
   members of
   the group's server map, without being aware that their
   packets will node.  A single node may register multiple times if it supports
   multiple layer 3 protocols. The CMIs allocated by the MARS for each
   such registration may or may not be going directly to the multicast group's members.

6.2.2   MARS_MSERV and MARS_UNSERV messages.

   MARS_MSERV and MARS_UNSERV are identical to the same.

   The retransmitted registration MARS_JOIN message.
   An MCS uses must NOT be sent on
   ClusterControlVC.  If a MARS_MSERV with cluster member issues a <min,max> pair of <X,X> to specify
   the multicast group X that deregistration
   MARS_LEAVE it too is willing to support. A single group
   MARS_UNSERV indicates the group retransmitted privately.

   Non-registration MARS_JOIN and MARS_LEAVE messages are ignored if
   they arrive from a node that the MCS is no longer willing to
   support. not registered as a cluster member.

   Except as described in section 6.2.4, after performing any required
   database updates non-registration MARS_JOIN and MARS_LEAVE messages
   are retransmitted on ClusterControlVC.  The operation code for MARS_MSERV following fields are
   modified just prior to retransmission:

      ar$flags.copy is 13 (decimal), and
   MARS_UNSERV set to 1.

      ar$msn is 17 (decimal).

   When operating on specific groups ar$flags.register MUST be zero.

   When an MCS issues a MARS_MSERV for a specific group set to the message MUST
   be dropped current Cluster Sequence Number for
      ClusterControlVC (Section 5.1.4.2).

   The MARS retransmits MARS_JOIN and ignored MARS_LEAVE messages even if they
   resulted in no change to the source has not already registered with
   the MARS as a multicast server (section 6.2.3).  Otherwise, database.

   MARS_JOIN or MARS_LEAVE messages MUST arrive at the MARS
   adds the new ATM address with
   ar$flags.copy set to 0, otherwise the server map for the specified group,
   possibly constructing a new server map if this message is the first MCS silently ignored.
   All outgoing MARS_JOIN or MARS_LEAVE messages have ar$flags.copy set
   to 1.

   ar$flags.layer3grp (section 5.3) MUST be ignored (and treated as
   reset) for
   the MARS_JOINs specifying more than a single group.

   When an MCS issues If a MARS_UNSERV the MARS removes its ATM address
   from the server maps for each specified group, deleting any server
   maps
   MARS_JOIN is received that end up being null after the operation.

   Both of these messages are sent to contains more than one <min,max> pair, the
   MARS over a point to point VC
   (between MCS and MARS). After processing, they are retransmitted on
   ServerControlVC to allow other MCSs to note MUST ignore the new node.

   The operation code is then changed to MARS_JOIN or MARS_LEAVE second and subsequent pairs.



Armitage                Expires March 20th, April 24th, 1996                [Page 40] 43]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   respectively, and another copy


   An additional IPv4 specific behaviour exists - if a node issues a
   MARS_LEAVE for address "224.0.0.1" (the 'all systems' group) it is
   assumed to have ceased multicast support completely. All references
   to this node MUST be eliminated from any other IPv4 groups it is a
   member of in the message is also transmitted on
   ClusterControlVC. This fools database. However, the cluster members into thinking endpoint is NOT released as a new
   leaf node as been added to (or dropped from) the group specified. In from ClusterControlVC (this only occurs upon receipt of a
   deregistration MARS_LEAVE).

   If the retransmitted MARS_JOIN/LEAVE ar$flags.layer3grp MUST be zero,
   ar$flags.copy MARS receives a deregistration MARS_LEAVE (described in
   section 5.2.3) that member's ATM address MUST be one, removed from all
   groups for which it may have joined, dropped from ClusterControlVC,
   and ar$flags.register MUST be zero.

   The the CMI released.

   If the MARS retransmits redundant MARS_MSERV receives an ERR_L_RELEASE on ClusterControlVC indicating
   that a cluster member has disconnected, that member's ATM address
   MUST be removed from all groups for which it may have joined, and MARS_UNSERV messages
   onto ServerControlVC, generates the appropriate MARS_JOIN or
   MARS_LEAVE messages
   CMI released.

6.1.3  Generating MARS_REDIRECT_MAP.

   A MARS_REDIRECT_MAP message (described in section 5.4.3) MUST be
   regularly transmitted on ClusterControlVC, but takes no further action. ClusterControlVC.  It is assumed RECOMMENDED that
   this occur every 1 minute, and it MUST occur at least one MCS will have MARS_MSERV'ed a group
   before the first cluster member joins it. every 2
   minutes. If a MARS_MSERV arrives for
   a group that the MARS has a non-null host map but no server map the default
   response knowledge of other backup MARSs serving
   the MARS will be to silently drop the MARS_MSERV without
   any further action. The MCS attempting to support cluster, it MUST include its own address as the group will
   eventually flag an error after repeated MARS_MSERVs fail.

   The last or only MCS for a group MAY choose to issue a MARS_UNSERV
   while the group still has members. When entry in the MARS_UNSERV is processed
   by
   MARS_REDIRECT_MAP message (in addition to filling in the source
   address fields).

   The design and use of backup MARS entities is beyond the 'server map' scope of
   this document, and will be deleted. When the associated
   MARS_LEAVE covered in future work.

6.1.4  Cluster Sequence Numbers.

   The Cluster Sequence Number (CSN) is issued on ClusterControlVC, all cluster members with a
   VC open to the MCS for that group will close down the VC (in
   accordance with described in section 5.1.4, since the MCS was their only leaf
   node). When and
   is carried in the ar$msn field of MARS messages being sent to cluster
   members subsequently find they need to transmit
   packets to the group, they will begin again with (either out ClusterControlVC or on an individual VC).  The
   MARS increments the
   MARS_REQUEST/MARS_MULTI sequence to establish CSN every time a new VC. Since message is sent on
   ClusterControlVC.  The current CSN is copied into the ar$msn field of
   MARS will have deleted the server map, this will result in the host
   map messages being return, and the group reverts sent to being supported by cluster members, whether out
   ClusterControlVC or on a VC
   mesh. private VC.

   A clean mechanism for MARS should be carefully designed to minimise the reverse process - transitioning a group
   from a VC mesh possibility of
   the CSN jumping unecessarily. Under normal operation only cluster
   members affected by transient link problems will miss CSN updates and
   be forced to MCS supported while revalidate. If the group is active - is a
   subject MARS itself glitches, it will be
   innundated with requests for further study.

6.2.3  Registering a Multicast Server (MCS).

   Section 5.2.3 describes how endpoints register period as every cluster members,
   and hence get added as leaf nodes member
   attempts to ClusterControlVC. The same
   approach revalidate.




Armitage                Expires April 24th, 1996                [Page 44]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   Calculations on the CSN MUST be performed as unsigned 32 bit
   arithmetic.

   One implication of this mechanism is used to register endpoints that intend to provide MCS
   support.

   Registration with the MARS occurs when an endpoint issues a
   MARS_MSERV with ar$flags.register set to one.  Upon registration the
   endpoint is added as a leaf node to ServerControlVC, should serialize
   its processing of 'simultaneous' MARS_REQUEST, MARS_JOIN and
   MARS_LEAVE messages. Join and Leave operations should be queued
   within the
   MARS_MSERV is returned to the MCS privately.

   The MCS retransmits this MARS_MSERV MARS along with MARS_REQUESTS, and not processed until it confirms that all
   the MARS
   has received it (by receiving reply packets of a copy back, in preceeding MARS_REQUEST have been transmitted.
   The transmission of MARS_REDIRECT_MAP should also be similarly
   queued.

   (The regular transmission of MARS_REDIRECT_MAP serves a secondary
   purpose of allowing cluster members to track the CSN, even if they
   miss an analogous way earlier MARS_JOIN or MARS_LEAVE.)

6.2   MARS interface to Multicast Servers (MCS).

   When the



Armitage                Expires March 20th, 1996                [Page 41]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


   mechanism MARS returns the actual addresses of group members, the
   endpoint behaviour described in section 5.2.2 for reliably transmitting
   MARS_JOINs).

   The ar$cmi field 5 results in MARS_MSERVs MUST be set to zero all groups being
   supported by both MCS and
   MARS.

   An MCS may also choose meshes of point to de-register, using a MARS_UNSERV with
   ar$flags.register set multipoint VCs. However, when MCSs
   register to one. When this occurs support particular layer 3 multicast groups the MARS MUST remove
   all references
   modifies its use of various MARS messages to that fool endpoints into
   using the MCS in all servermaps instead.

   The following MARS messages are associated with interaction between
   the
   protocol (ar$pro) specified MARS and MCSs.

      3   MARS_MSERV
      7   MARS_UNSERV
      8   MARS_SJOIN
      9   MARS_SLEAVE

   The following MARS messages are treated in the MARS_UNSERV.

   Note that multiple logical MCSs may share the same physical ATM
   interface, provided that each MCS uses a separate ATM address (e.g. a slightly different SEL field in the NSAP format address). In fact, an
   manner when MCSs have registered to support certain group addresses:

      1   MARS_REQUEST
      4   MARS_JOIN
      5   MARS_LEAVE

   A MARS must keep two sets of mappings for each layer 3 group using
   MCS may
   share support.  The original {layer 3 address, ATM.1, ATM.2, ... ATM.n}
   mapping (now termed the ATM interface of a node that 'host map', although it includes routers) is also a cluster member
   (either host or router), provided each logical entity has
   augmented by a different parallel {layer 3 address, server.1, server.2, ....
   server.K} mapping (the 'server map'). It is assumed that no ATM address.

   A MARS MUST be capable of handling a multi-entry servermap. However,
   addresses appear in both the possible use of server and host maps for the same
   multicast group. Typically K will be 1, but it will be larger if
   multiple MCSs registering are configured to support the same
   group is a subject for further study. In the absence of an MCS
   synchronisation protocol a system administrator MUST NOT allow more
   than one logical MCS to register for a given group.

6.2.4   Modified response to MARS_JOIN and MARS_LEAVE.




Armitage                Expires April 24th, 1996                [Page 45]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   The existence of MARS also maintains a point to multipoint VC out to any MCSs supporting some groups but not others requires
   registered with it, called ServerControlVC (section 6.2.3). This
   serves an analogous role to ClusterControlVC, allowing the MARS to modify
   update the MCSs with group membership changes as they occur. A MARS
   MUST also send its distribution of single regular MARS_REDIRECT_MAP transmissions on both
   ServerControlVC and block join/leave
   updates ClusterControlVC.

6.2.1   Response to cluster members. The MARS also adds two new messages -
   MARS_SJOIN and MARS_SLEAVE - for communicating group changes to MCSs
   over ServerControlVC.

   The MARS_SJOIN and MARS_SLEAVE messages are identical to MARS_JOIN,
   with operation codes 18 and 19 (decimal) respectively. a MARS_REQUEST if MCS is registered.

   When the MARS receives a cluster member issues MARS_JOIN or MARS_LEAVE MARS_REQUEST for an address that has both
   host and server maps it generates a single
   group, response based on the MARS checks to see if identity of
   the group has an associated server
   map. request's source. If the specified group does not have requestor is a member of the server map
   for the requested group then the MARS
   simply retransmits returns the MARS_JOIN contents of the
   host map in a sequence of one or MARS_LEAVE on ClusterControlVC.

   However, more MARS_MULTIs.  Otherwise, if the
   source is a valid cluster member, the MARS returns the contents of
   the server map exists for the group in a new set sequence of actions
   are taken.

      A copy one or more MARS_MULTIs.  If the
   source is neither a cluster member, nor a member of the MARS_JOIN/LEAVE server map
   for the group, the request is made with type MARS_SJOIN or
      MARS_SLEAVE as appropriate, dropped and transmitted on ServerControlVC.
      This allows ignored.

   Servers use the MCS(s) supporting host map to establish a basic distribution VC for the group
   group. Cluster members will establish outgoing multipoint VCs to note
   members of the new member
      and update group's server map, without being aware that their data VCs.




Armitage                Expires March 20th, 1996                [Page 42]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


      The original message is transmitted back
   packets will not be going directly to the source cluster
      member unchanged, using the VC it arrived on rather than
      ClusterControlVC.  The ar$flags.punched field MUST be reset to 0
      in this message.

   (Section 5.2.2 requires cluster members have a mechanism to confirm
   the reception of their message by the MARS. For mesh supported
   groups, using ClusterControlVC serves dual purpose of providing this
   confirmation multicast group's members.

6.2.2   MARS_MSERV and distributing group update information. When a group
   is MCS supported, there is no reason for all cluster members to
   process null join/leave messages on ClusterControlVC, so they MARS_UNSERV messages.

   MARS_MSERV and MARS_UNSERV are
   sent back on identical to the private VC between cluster member and MARS.)

   Receipt of a block MARS_JOIN (e.g. from message.
   An MCS uses a router coming on-line) or
   MARS_LEAVE requires MARS_MSERV with a more complex response. The single <min,max>
   block may simultaneously cover mesh supported and MCS supported
   groups.  However, cluster members only need to be informed pair of the
   mesh supported groups that the endpoint has joined. Only the MCSs
   need <X,X> to know if specify
   the endpoint is joining any MCS supported groups.

   The solution multicast group X that it is willing to modify support. A single group
   MARS_UNSERV indicates the MARS_JOIN or MARS_LEAVE group that the MCS is
   retransmitted on ClusterControlVC. no longer willing to
   support.  The following action is taken:

      A copy of the MARS_JOIN/LEAVE operation code for MARS_MSERV is made with type MARS_SJOIN or
      MARS_SLEAVE as appropriate, 3 (decimal), and transmitted
   MARS_UNSERV is 7 (decimal).

   When operating on ServerControlVC.
      This allows the MCS(s) supporting the specific groups ar$flags.register MUST be zero.

   When an MCS issues a MARS_MSERV for a specific group to note the membership
      change message MUST
   be dropped and update their outgoing point to multipoint VCs.

      The <min,max> block supplied in ignored if the original MARS_JOIN/LEAVE is
      replaced source has not already registered with
   the MARS as a 'hole punched' set of zero or more <min,max>
      pairs.  The 'hole punched' set of <min,max> pairs covers multicast server (section 6.2.3).  Otherwise, the
      entire MARS
   adds the new ATM address range specified by to the original <min,max> pair, but
      excludes those addresses/groups supported by MCSs.

      If no 'holes' were punched in server map for the specified block, the original
      MARS_JOIN/LEAVE is re-transmitted out on ClusterControlVC
      unchanged.  Otherwise the following occurs:

         The original MARS_JOIN/LEAVE group,
   possibly constructing a new server map if this is transmitted back to the source
         cluster member unchanged, using the VC it arrived on. The
         ar$flags.punched field MUST be reset to 0 in this message.

         If first MCS for
   the hole-punched set contains 1 or more <min,max> pair, group.

   When an MCS issues a
         copy of MARS_UNSERV the original MARS_JOIN/LEAVE is transmitted on
         ClusterControlVC, carrying MARS removes its ATM address
   from the new <min,max> list. The
         ar$flags.punched field MUST be set to 1 in this message.

      The ar$flags.punched field is set server maps for each specified group, deleting any server
   maps that end up being null after the operation.

   Both of these messages are sent to ensure the hole-punched copy MARS over a point to point VC
   (between MCS and MARS). After processing, they are retransmitted on



Armitage                Expires March 20th, April 24th, 1996                [Page 43] 46]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


      is ignored by the message's source when trying


   ServerControlVC to match received
      MARS_JOIN/LEAVE messages with ones previously sent (section
      5.2.2).

   (Appendix A discusses some algorithms for 'hole punching'.)

   It is assumed that allow other MCSs use the MARS_SJOINs and MARS_SLEAVEs to
   update their own VCs out to note the actual group's members.

   ar$flags.layer3grp new node.

   The operation code is copied over into then changed to MARS_JOIN or MARS_LEAVE
   respectively, and another copy of the messages message is also transmitted by on
   ClusterControlVC. This fools the MARS. cluster members into thinking a new
   leaf node as been added to (or dropped from) the group specified. In
   the retransmitted MARS_JOIN/LEAVE ar$flags.layer3grp MUST be zero,
   ar$flags.copy MUST be set to one.

6.2.5  Sequence numbers for ServerControlVC traffic.

   In an analogous fashion to one, and ar$flags.register MUST be zero.

   The MARS retransmits redundant MARS_MSERV and MARS_UNSERV messages
   onto ServerControlVC, generates the Cluster Sequence Number, appropriate MARS_JOIN or
   MARS_LEAVE messages on ClusterControlVC, but takes no further action.

   It is assumed that at least one MCS will have MARS_MSERV'ed a group
   before the MARS
   keeps first cluster member joins it. If a Server Sequence Number (SSN) that is incremented MARS_MSERV arrives for every
   transmission on ServerControlVC. The current value of the SSN is
   inserted into
   a group that has a non-null host map but no server map the ar$msn field default
   response of every message the MARS issues that
   it believes is destined for an MCS. This includes MARS_MULTIs that
   are being returned in response will be to a MARS_REQUEST from an MCS, and
   MARS_REDIRECT_MAP being sent on ServerControlVC. silently drop the MARS_MSERV without
   any further action. The MCS must check attempting to support the MARS_REQUESTs source, and if it is a registered group will
   eventually flag an error after repeated MARS_MSERVs fail.

   The last or only MCS for a group MAY choose to issue a MARS_UNSERV
   while the SSN group still has members. When the MARS_UNSERV is
   copied into processed
   by the ar$msn field, otherwise the CSN is copied into MARS the
   ar$msn field.

   MCSs are expected to track and use 'server map' will be deleted. When the SSNs in an analogous manner associated
   MARS_LEAVE is issued on ClusterControlVC, all cluster members with a
   VC open to the way endpoints use the CSN in section 5.1 (to trigger revalidation
   of MCS for that group membership information).

   A MARS should be carefully designed to minimise will close down the possibility of VC (in
   accordance with section 5.1.4, since the SSN jumping unecessarily. Under normal operation MCS was their only MCSs that
   are affected by transient link problems will miss ar$msn updates and
   be forced leaf
   node). When cluster members subsequently find they need to transmit
   packets to revalidate. If the MARS itself glitches it group, they will be
   innundated begin again with requests for a period as every MCS attempts to
   revalidate.

6.3 Why global sequence numbers?

   The CSN and SSN are global within the context of
   MARS_REQUEST/MARS_MULTI sequence to establish a given protocol
   (e.g. IP).  They count ClusterControlVC new VC. Since the
   MARS will have deleted the server map, this will result in the host
   map being return, and ServerControlVC activity
   without reference to the multicast group(s) involved.  This may be
   perceived as group reverts to being supported by a limitation, because there is no way VC
   mesh.

   A clean mechanism for cluster
   members or multicast servers to isolate exactly which multicast the reverse process - transitioning a group
   they may have missed an update for. An alternative was
   from a VC mesh to try and
   provide MCS supported while the group is active - is a per-group sequence number.

   Unfortunately per-group sequence numbers are not practical.
   subject for further study.

6.2.3  Registering a Multicast Server (MCS).

   Section 5.2.3 describes how endpoints register as cluster members,
   and hence get added as leaf nodes to ClusterControlVC. The
   current mechanism allows sequence information same
   approach is used to be piggy-backed onto register endpoints that intend to provide MCS
   support.

   Registration with the MARS occurs when an endpoint issues a
   MARS_MSERV with ar$flags.register set to one.  Upon registration the
   endpoint is added as a leaf node to ServerControlVC, and the
   MARS_MSERV is returned to the MCS privately.



Armitage                Expires March 20th, April 24th, 1996                [Page 44] 47]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   MARS messages already in transit for other reasons.


   The ability to
   specify blocks of multicast addresses with a single MARS_JOIN or
   MARS_LEAVE means MCS retransmits this MARS_MSERV until it confirms that the MARS
   has received it (by receiving a single message can refer copy back, in an analogous way to membership change the
   mechanism described in section 5.2.2 for multiple groups simultaneously. A single ar$msn reliably transmitting
   MARS_JOINs).

   The ar$cmi field cannot
   provide meaningful information about each group's sequence.  Multiple
   ar$msn fields would have been unwieldy.

   Any MARS or cluster member that supports different protocols in MARS_MSERVs MUST
   keep separate mapping tables and sequence numbers for each protocol.

6.4 Redundant/Backup MARS Architectures.

   If backup MARSs exist for a given cluster then mechanisms are needed be set to ensure consistency between their mapping tables zero by both MCS and those of the
   active, current
   MARS.

   (Cluster members will consider backup MARSs

   An MCS may also choose to exist if they have
   been configured with de-register, using a table of MARS addresses, or the regular
   MARS_REDIRECT_MAP messages contain a list of 2 or more addresses.)

   The definition of an MARS-synchronization protocol is beyond the
   current scope of MARS_UNSERV with
   ar$flags.register set to one. When this document, and is expected occurs the MARS MUST remove
   all references to be that MCS in all servermaps associated with the subject of
   further research work.  However,
   protocol (ar$pro) specified in the following observations MARS_UNSERV.

   Note that multiple logical MCSs may be
   made:

      The MARS_REDIRECT_MAP message exist, enabling one MARS to force
      endpoints to move to another MARS share the same physical ATM
   interface, provided that each MCS uses a separate ATM address (e.g. a
   different SEL field in the aftermath NSAP format address). In fact, an MCS may
   share the ATM interface of a node that is also a cluster member
   (either host or router), provided each logical entity has a different
   ATM address.

   A MARS
      failure, the chosen backup MARS will eventually wish to hand
      control MUST be capable of handling a multi-entry servermap. However,
   the cluster over to the main MARS when it is
      functioning properly again).

      Cluster members and possible use of multiple MCSs do not need registering to start up with knowledge of
      more than one MARS, provided that MARS correctly issues
      MARS_REDIRECT_MAP messages with support the full list of MARSs for that
      cluster.

   Any mechanism same
   group is a subject for synchronising backup MARSs (and coping with further study. In the
   aftermath absence of MARS failures) should be compatible with the cluster
   member behaviour described in this document.

7.   How an MCS utilises
   synchronisation protocol a MARS.

   When an system administrator MUST NOT allow more
   than one logical MCS supports a multicast to register for a given group.

6.2.4   Modified response to MARS_JOIN and MARS_LEAVE.

   The existence of MCSs supporting some groups but not others requires
   the MARS to modify its distribution of single and block join/leave
   updates to cluster members. The MARS also adds two new messages -
   MARS_SJOIN and MARS_SLEAVE - for communicating group it acts as changes to MCSs
   over ServerControlVC.

   The MARS_SJOIN and MARS_SLEAVE messages are identical to MARS_JOIN,
   with operation codes 18 and 19 (decimal) respectively.

   When a proxy cluster
   endpoint member issues MARS_JOIN or MARS_LEAVE for a single
   group, the senders MARS checks to see if the group. It also behaves in group has an
   analogous manner to associated server
   map. If the specified group does not have a sender, managing server map the MARS
   simply retransmits the MARS_JOIN or MARS_LEAVE on ClusterControlVC.

   However, if a single outgoing point to
   multipoint VC to server map exists for the real group members.

   Detailed description a new set of possible MCS architectures actions
   are beyond taken.

      A copy of the MARS_JOIN/LEAVE is made with type MARS_SJOIN or
      MARS_SLEAVE as appropriate, and transmitted on ServerControlVC.
      This allows the MCS(s) supporting the group to note the new member



Armitage                Expires March 20th, April 24th, 1996                [Page 45] 48]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   scope of this document. This section will outline the main issues.

7.1   Association with a particular Layer 3 group.

   When an MCS issues a MARS_MSERV it forces all senders to the
   specified layer 3 group to terminate


      and update their VCs on the supplied source
   ATM address. data VCs.

      The simplest MCS architecture involves taking incoming AAL_SDUs and
   simply flipping them original message is transmitted back out a single point to multipoint VC. Such
   an MCS cannot support more than one group at once, as the source cluster
      member unchanged, using the VC it has no way arrived on rather than
      ClusterControlVC.  The ar$flags.punched field MUST be reset to differentiate between traffic destined for different groups.
   Using 0
      in this architecture, message.

   (Section 5.2.2 requires cluster members have a physical node would provide MCS support
   for multiple groups mechanism to confirm
   the reception of their message by creating multiple logical instances the MARS. For mesh supported
   groups, using ClusterControlVC serves dual purpose of providing this
   confirmation and distributing group update information. When a group
   is MCS supported, there is no reason for all cluster members to
   process null join/leave messages on ClusterControlVC, so they are
   sent back on the
   MCS, each with different ATM Addresses private VC between cluster member and MARS.)

   Receipt of a block MARS_JOIN (e.g. from a router coming on-line) or
   MARS_LEAVE requires a different SEL value in
   the node's NSAPA).

   A slightly more complex approach would response. The single <min,max>
   block may simultaneously cover mesh supported and MCS supported
   groups.  However, cluster members only need to be informed of the
   mesh supported groups that the endpoint has joined. Only the MCSs
   need to add minimal layer 3
   specific processing into know if the MCS. This would look inside endpoint is joining any MCS supported groups.

   The solution is to modify the received
   AAL_SDUs and determine which layer 3 group they are destined for. MARS_JOIN or MARS_LEAVE that is
   retransmitted on ClusterControlVC. The following action is taken:

      A
   single instance copy of such an MCS might register its ATM Address with the MARS for multiple layer 3 groups, MARS_JOIN/LEAVE is made with type MARS_SJOIN or
      MARS_SLEAVE as appropriate, and manage multiple independent transmitted on ServerControlVC.
      This allows the MCS(s) supporting the group to note the membership
      change and update their outgoing point to multipoint VCs (one for each group).

   When an MCS starts up it MUST register with VCs.

      Before transmission on the MARS as described in
   section 6.2.3, identifying ClusterControlVC, the protocol it supports original
      MARS_JOIN/LEAVE then has its <min,max> block replaced with the ar$pro
   field a 'hole
      punched' set of zero or more <min,max> pairs.  The 'hole punched'
      set of <min,max> pairs covers the MARS_MSERV. This also applies to logical MCSs, even if
   they share entire address range specified
      by the same physical ATM interface. This original <min,max> pair, but excludes those
      addresses/groups supported by MCSs.

      If no 'holes' were punched in the specified block, the original
      MARS_JOIN/LEAVE is important so that re-transmitted out on ClusterControlVC
      unchanged.  Otherwise the MARS can react following occurs:

         The original MARS_JOIN/LEAVE is transmitted back to the loss of an MCS when it drops off
   ServControlVC. (One consequence is that 'simple' MCS architectures
   end up with one ServerControlVC member per group.  MCSs with layer 3
   specific processing may support multiple groups while still only
   registering as one source
         cluster member of ServerControlVC.)

   An MCS MUST NOT share unchanged, using the same ATM address as a cluster member,
   although VC it may share the same physical ATM interface.

7.2   Termination of incoming VCs.

   An MCS arrived on. The
         ar$flags.punched field MUST terminate unidirectional VCs be reset to 0 in this message.

         If the same manner as hole-punched set contains 1 or more <min,max> pair, a
   cluster member.  (e.g. terminate on an LLC entity when LLC/SNAP
   encapsulation is used, as described in RFC 1755 for unicast
   endpoints.)

7.3   Management
         copy of outgoing VC.

   An MCS MUST establish and manage its outgoing point to multipoint VC
   as a cluster member does (section 5.1). the original MARS_JOIN/LEAVE is transmitted on
         ClusterControlVC, carrying the new <min,max> list. The



Armitage                Expires March 20th, April 24th, 1996                [Page 46] 49]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   MARS_REQUEST


         ar$flags.punched field MUST be set to 1 in this message.

      The ar$flags.punched field is used by the MCS set to establish ensure the initial leaf nodes
   for hole-punched copy
      is ignored by the MCS's outgoing point message's source when trying to multipoint VC. After the VC match received
      MARS_JOIN/LEAVE messages with ones previously sent (section
      5.2.2).

   (Appendix A discusses some algorithms for 'hole punching'.)

   It is
   established, assumed that MCSs use the MCS reacts to MARS_SJOINs and MARS_SLEAVEs in the
   same way a cluster member reacts to MARS_JOINs and MARS_LEAVEs.

   The MCS tracks
   update their own VCs out to the Server Sequence Number from actual group's members.

   ar$flags.layer3grp is copied over into the ar$msn fields of messages from transmitted by
   the MARS, and revalidates its outgoing point to
   multipoint VC(s) when a sequence number jump occurs.

7.4   Use of a backup MARS.

   The MCS uses the same approach ar$flags.copy MUST be set to backup MARSs as a cluster member
   (section 5.4), tracking MARS_REDIRECT_MAP messages on
   ServerControlVC.


8.   Support for IP multicast routers.

   Multicast routers are required one.

6.2.5  Sequence numbers for ServerControlVC traffic.

   In an analogous fashion to the propagation of multicast
   traffic beyond Cluster Sequence Number, the constraints of MARS
   keeps a single cluster (inter-cluster
   traffic).  (There Server Sequence Number (SSN) that is a sense in which they are multicast servers
   acting at the next higher layer, with clusters, rather than
   individual endpoints, as their abstract sources and destinations.)

   Multicast routers typically participate in higher layer multicast
   routing algorithms and policies that are beyond the scope incremented for every
   transmission on ServerControlVC. The current value of this
   memo (e.g. DVMRP [5] in the IPv4 environment).

   It SSN is assumed that the multicast routers will be implemented over
   inserted into the
   same sort ar$msn field of IP/ATM interface that a multicast host would use.  Their
   IP/ATM interfaces will will register with every message the MARS as a cluster
   members, joining and leaving multicast groups as necessary. As noted issues that
   it believes is destined for an MCS. This includes MARS_MULTIs that
   are being returned in section 5, multiple logical 'endpoints' may be implemented over a
   single physical ATM interface. Routers use this approach response to provide
   interfaces into each clusters they will be routing between.

   The rest of this section will assume a simple IPv4 scenario where MARS_REQUEST from an MCS, and
   MARS_REDIRECT_MAP being sent on ServerControlVC.  The MCS must check
   the
   scope of a cluster has been limited to MARS_REQUESTs source, and if it is a particular LIS that registered MCS the SSN is part
   of an overlaid IP network. Not all members of
   copied into the LIS are necessarily
   registered cluster members (you may have unicast-only hosts in ar$msn field, otherwise the
   LIS).

8.1    Forwarding CSN is copied into a Cluster.

   If the multicast router needs to transmit a packet
   ar$msn field.

   MCSs are expected to a group within track and use the cluster its IP/ATM interface opens a VC SSNs in the same an analogous manner as a to
   the way endpoints use the CSN in section 5.1 (to trigger revalidation
   of group membership information).

   A MARS should be carefully designed to minimise the possibility of
   the SSN jumping unecessarily. Under normal host would. Once a VC is open, operation only MCSs that
   are affected by transient link problems will miss ar$msn updates and
   be forced to revalidate. If the router watches MARS itself glitches it will be
   innundated with requests for
   MARS_JOIN a period as every MCS attempts to
   revalidate.

6.3 Why global sequence numbers?

   The CSN and MARS_LEAVE messages SSN are global within the context of a given protocol
   (e.g. IP).  They count ClusterControlVC and responds ServerControlVC activity
   without reference to them the multicast group(s) involved.  This may be
   perceived as a normal limitation, because there is no way for cluster
   members or multicast servers to isolate exactly which multicast group
   they may have missed an update for. An alternative was to try and
   provide a per-group sequence number.



Armitage                Expires March 20th, April 24th, 1996                [Page 47] 50]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   host would.

   The multicast router's transmit side MUST implement inactivity timers
   to shut down idle outgoing VCs, as for normal hosts.

   As with normal host, the multicast router does


   Unfortunately per-group sequence numbers are not need practical. The
   current mechanism allows sequence information to be a
   member of a group it is sending to.

8.2    Joining piggy-backed onto
   MARS messages already in 'promiscuous' mode.

   Once registered and initialised, the simplest model of IPv4 multicast
   router operation is transit for it other reasons. The ability to issue
   specify blocks of multicast addresses with a single MARS_JOIN encompassing the
   entire Class D address space.  In effect it becomes 'promiscuous', as
   it will be or
   MARS_LEAVE means that a leaf node single message can refer to all present membership change
   for multiple groups simultaneously. A single ar$msn field cannot
   provide meaningful information about each group's sequence.  Multiple
   ar$msn fields would have been unwieldy.

   Any MARS or cluster member that supports different protocols MUST
   keep separate mapping tables and future multipoint VCs
   established sequence numbers for each protocol.

6.4 Redundant/Backup MARS Architectures.

   If backup MARSs exist for a given cluster then mechanisms are needed
   to IPv4 groups on ensure consistency between their mapping tables and those of the cluster.

   How a router chooses which groups
   active, current MARS.

   (Cluster members will consider backup MARSs to propagate outside exist if they have
   been configured with a table of MARS addresses, or the cluster regular
   MARS_REDIRECT_MAP messages contain a list of 2 or more addresses.)

   The definition of an MARS-synchronization protocol is beyond the
   current scope of this document.

   Consistent with RFC 1112, IP multicast routers may retain the use of
   IGMP Query document, and IGMP Report messages is expected to ascertain group membership.
   However, certain optimisations are possible, and are described in
   section 8.5.

8.3    Forwarding across be the cluster.

   Under some circumstances subject of
   further research work.  However, the cluster following observations may simply be
   made:

      The MARS_REDIRECT_MAP message exist, enabling one MARS to force
      endpoints to move to another hop
   between IP subnets that have participants MARS (e.g. in a multicast group.

      [LAN.1] ----- IPmcR.1 -- [cluster/LIS] -- IPmcR.2 ----- [LAN.2]

   LAN.1 and LAN.2 are subnets (such as Ethernet) with attached hosts
   that are members the aftermath of group X.

   IPmcR.1 and IPmcR.2 are multicast routers with interfaces to a MARS
      failure, the LIS.

   A traditional solution would be chosen backup MARS will eventually wish to treat the LIS as a unicast subnet,
   and use tunneling routers. However, this would not allow hosts on hand
      control of the
   LIS cluster over to participate in the cross-LIS traffic.

   Assume IPmcR.1 is receiving packets promiscuously on its LAN.1
   interface. Assume further main MARS when it is configured to propagate multicast
   traffic
      functioning properly again).

      Cluster members and MCSs do not need to all attached interfaces. In this case start up with knowledge of
      more than one MARS, provided that means MARS correctly issues
      MARS_REDIRECT_MAP messages with the LIS.

   When a packet full list of MARSs for group X arrives on its LAN.1 interface, IPmcR.1
   simply sends that
      cluster.

   Any mechanism for synchronising backup MARSs (and coping with the packet to group X on
   aftermath of MARS failures) should be compatible with the LIS interface cluster
   member behaviour described in this document.

7.   How an MCS utilises a MARS.

   When an MCS supports a multicast group it acts as a normal
   host would (Issuing MARS_REQUEST proxy cluster
   endpoint for group X, creating the VC,
   sending senders to the packet). group. It also behaves in an
   analogous manner to a sender, managing a single outgoing point to
   multipoint VC to the real group members.



Armitage                Expires March 20th, April 24th, 1996                [Page 48] 51]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   Assuming IPmcR.2 initialised itself with the MARS as a member


   Detailed description of possible MCS architectures are beyond the
   entire Class D space, it will have been returned as a member
   scope of X
   even if no other nodes on this document. This section will outline the LIS were members. All packets for main issues.

7.1   Association with a particular Layer 3 group.

   When an MCS issues a MARS_MSERV it forces all senders to the
   specified layer 3 group
   X received on IPmcR.2's LIS interface may be retransmitted to terminate their VCs on LAN.2.

   If IPmcR.1 is similarly initialised the reverse process will apply
   for multicast traffic from LAN.2 to LAN.1, for any multicast group. supplied source
   ATM address.

   The benefit of this scenario is that cluster members within the LIS
   may also join simplest MCS architecture involves taking incoming AAL_SDUs and leave
   simply flipping them back out a single point to multipoint VC. Such
   an MCS cannot support more than one group X at anytime.

8.4   Joining in 'semi-promiscuous' mode.

   Both unicast and multicast IP routers have once, as it has no way
   to differentiate between traffic destined for different groups.
   Using this architecture, a common problem -
   limitations on the number physical node would provide MCS support
   for multiple groups by creating multiple logical instances of AAL contexts available at their ATM
   interfaces.  Being 'promiscuous' in the RFC 1112 sense means that for
   every M hosts sending to N groups, a multicast router's
   MCS, each with different ATM interface
   will have M*N incoming reassembly engines tied up.

   It is not hard to envisage situations where Addresses (e.g. a number of multicast
   groups are active within the LIS but are not required to be
   propagated beyond different SEL value in
   the LIS itself. An example might node's NSAPA).

   A slightly more complex approach would be a distributed
   simulation system specifically designed to use the high speed IP/ATM
   environment. There may be no practical way its traffic could be
   utilised on 'the other side' of add minimal layer 3
   specific processing into the multicast router, yet under MCS. This would look inside the
   conventional scheme received
   AAL_SDUs and determine which layer 3 group they are destined for. A
   single instance of such an MCS might register its ATM Address with
   the router would have to be a leaf MARS for multiple layer 3 groups, and manage multiple independent
   outgoing point to multipoint VCs (one for each
   participating host anyway.

   As this problem occurs below the IP layer, group).

   When an MCS starts up it is worth noting that
   'scoping' mechanisms at MUST register with the IP multicast routing level do not provide
   a solution. An IP level scope would still result MARS as described in
   section 6.2.3, identifying the router's ATM
   interface receiving traffic on protocol it supports with the scoped groups, only ar$pro
   field of the MARS_MSERV. This also applies to drop it.

   In this situation logical MCSs, even if
   they share the network administrator might configure their
   multicast routers same physical ATM interface. This is important so that
   the MARS can react to exclude sections of the Class D address space loss of an MCS when issuing MARS_JOIN(s). Multicast groups it drops off
   ServControlVC. (One consequence is that will never be
   propagated beyond the cluster will not have 'simple' MCS architectures
   end up with one ServerControlVC member per group.  MCSs with layer 3
   specific processing may support multiple groups while still only
   registering as one member of ServerControlVC.)

   An MCS MUST NOT share the router listed same ATM address as a cluster member, and the router will never have to receive (and simply ignore)
   traffic from those groups.

   Another scenario involves the product M*N exceeding
   although it may share the capacity same physical ATM interface.

7.2   Termination of a
   single router's interface (especially if incoming VCs.

   An MCS MUST terminate unidirectional VCs in the same interface must also
   support a unicast IP router service).

   A network administrator may choose to add a second node, to function manner as a parallel IP multicast router. Each router would be configured to
   be 'promiscuous' over separate parts of the Class D address space,
   thus exposing themselves to only part of the VC load. This sharing
   would be completely transparent to IP hosts within the LIS.
   cluster member.  (e.g. terminate on an LLC entity when LLC/SNAP
   encapsulation is used, as described in RFC 1755 for unicast
   endpoints.)







Armitage                Expires March 20th, April 24th, 1996                [Page 49] 52]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   Restricted promiscuous mode does not break RFC 1112's use


7.3   Management of IGMP
   Report messages. If the router is configured outgoing VC.

   An MCS MUST establish and manage its outgoing point to serve multipoint VC
   as a given block
   of Class D addresses, it will receive the IGMP Report.  If the router cluster member does (section 5.1).

   MARS_REQUEST is not configured used by the MCS to support a given block, then establish the existence of an
   IGMP Report initial leaf nodes
   for a group in that block is irrelevant to the router.
   All routers are able MCS's outgoing point to track membership changes through multipoint VC. After the
   MARS_JOIN VC is
   established, the MCS reacts to MARS_SJOINs and MARS_LEAVE traffic anyway. (Section 8.5 discusses MARS_SLEAVEs in the
   same way a
   better alternative cluster member reacts to IGMP within a cluster.)

   Mechanisms MARS_JOINs and reasons for establishing these modes of operation are
   beyond MARS_LEAVEs.

   The MCS tracks the scope Server Sequence Number from the ar$msn fields of this document.

8.5   An alternative
   messages from the MARS, and revalidates its outgoing point to IGMP Queries.

   An unfortunate aspect of IGMP is that it assumes multicasting
   multipoint VC(s) when a sequence number jump occurs.

7.4   Use of IP
   packets is a cheap and trivial event at backup MARS.

   The MCS uses the link layer. As a
   consequence, regular IGMP Queries are multicasted by routers to group
   224.0.0.1. These queries are intended same approach to trigger IGMP Replies by backup MARSs as a cluster members that have layer 3 members of particular groups.

   The MARS_GROUPLIST_REQUEST and MARS_GROUPLIST_REPLY member
   (section 5.4), tracking MARS_REDIRECT_MAP messages were
   designed to allow on
   ServerControlVC.


8.   Support for IP multicast routers.

   Multicast routers to avoid actually transmitting IGMP Queries
   out into a cluster.

   Whenever are required for the router's forwarding engine wishes to transmit an IGMP
   query, a MARS_GROUPLIST_REQUEST can be sent to propagation of multicast
   traffic beyond the MARS instead. The
   resulting MARS_GROUPLIST_REPLY(s) (described constraints of a single cluster (inter-cluster
   traffic).  (There is a sense in section 5.3) from the
   MARS carry all which they are multicast servers
   acting at the information next higher layer, with clusters, rather than
   individual endpoints, as their abstract sources and destinations.)

   Multicast routers typically participate in higher layer multicast
   routing algorithms and policies that are beyond the router would have ascertained
   from IGMP replies. scope of this
   memo (e.g. DVMRP [5] in the IPv4 environment).

   It is RECOMMENDED assumed that the multicast routers utilise this MARS service to
   minimise IGMP traffic within the cluster.

   By default a MARS_GROUPLIST_REQUEST SHOULD specify will be implemented over the entire address
   space (e.g. <224.0.0.0, 239.255.255.255> in an IPv4 environment).
   However, routers serving part
   same sort of IP/ATM interface that a multicast host would use.  Their
   IP/ATM interfaces will will register with the address space (as described MARS as a cluster
   members, joining and leaving multicast groups as necessary. As noted
   in section 8.4) MAY choose 5, multiple logical 'endpoints' may be implemented over a
   single physical ATM interface. Routers use this approach to issue MARS_GROUPLIST_REQUESTs that specify
   only the subset of the address space provide
   interfaces into each clusters they are serving.

   (On the surface it would also seem useful for multicast routers to
   track MARS_JOINs and MARS_LEAVEs that arrive with ar$flags.layer3grp
   set. These might will be used in lieu routing between.

   The rest of IGMP Reports, to provide the
   router with timely indication that a new layer 3 group member exists
   within the cluster. However, this only works on VC mesh supported
   groups, and is therefore NOT recommended).

   Appendix B discusses less elegant mechanisms for reducing section will assume a simple IPv4 scenario where the impact
   scope of IGMP traffic within a cluster, on the assumption cluster has been limited to a particular LIS that is part
   of an overlaid IP network. Not all members of the IP/ATM LIS are necessarily
   registered cluster members (you may have unicast-only hosts in the
   LIS).





Armitage                Expires March 20th, April 24th, 1996                [Page 50] 53]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   interfaces


8.1    Forwarding into a Cluster.

   If the multicast router needs to transmit a packet to a group within
   the cluster are being used by un-optimised IP
   multicasting code.

8.6   CMIs across multiple interfaces.

   The Cluster Member ID is only unique within the Cluster managed by its IP/ATM interface opens a
   given MARS. On VC in the surface this might appear to leave us with same manner as a
   problem when
   normal host would. Once a multicast router VC is routing between two or more
   Clusters using a single physical ATM interface.  The open, the router will
   register with two or more MARSs, watches for
   MARS_JOIN and thereby acquire two or more
   independent CMI's. Given that each MARS has no reason MARS_LEAVE messages and responds to synchronise
   their CMI allocations, it is possible for them as a normal
   host in one cluster to
   have the same CMI has the router's interface to another Cluster. How
   does the router distinguish between its own reflected packets, and
   packets from that other host? would.

   The answer lies in the fact that routers (and hosts) actually multicast router's transmit side MUST implement logical IP/ATM interfaces over a single physical ATM
   interface. Each logical interface will have a unique ATM Address (eg.
   an NSAP with different SELector fields, one inactivity timers
   to shut down idle outgoing VCs, as for each logical
   interface).

   Each logical IP/ATM interface is configured normal hosts.

   As with normal host, the address multicast router does not need to be a
   member of a
   single MARS, attaches to only one cluster, group it is sending to.

8.2    Joining in 'promiscuous' mode.

   Once registered and so had only one CMI to
   worry about. Each of the MARSs that initialised, the simplest model of IPv4 multicast
   router operation is registered with
   will have been given a different ATM Address (corresponding for it to the
   different logical IP/ATM interfaces) in each registration MARS_JOIN.

   When hosts in issue a cluster add MARS_JOIN encompassing the router
   entire Class D address space.  In effect it becomes 'promiscuous', as
   it will be a leaf node, they'll
   specify the ATM Address of the appropriate logical IP/ATM interface node to all present and future multipoint VCs
   established to IPv4 groups on the cluster.

   How a router in the L_MULTI_ADD message. Thus, each logical IP/ATM
   interface will only have chooses which groups to check and filter on CMIs assigned by its
   own MARS.

   In essence propagate outside the cluster differentiation is achieved by ensuring that
   logical IP/ATM interfaces are assigned different ATM Addresses.


9.    Multiprotocol applications
   beyond the scope of this document.

   Consistent with RFC 1112, IP multicast routers may retain the MARS use of
   IGMP Query and MARS clients.

   A deliberate attempt has been made IGMP Report messages to describe the MARS ascertain group membership.
   However, certain optimisations are possible, and
   associated mechanisms are described in a manner independent of a specific higher
   layer protocol being run over
   section 8.5.

8.3    Forwarding across the ATM cloud. The immediate
   application of this document will cluster.

   Under some circumstances the cluster may simply be another hop
   between IP subnets that have participants in an IPv4 environment, a multicast group.

      [LAN.1] ----- IPmcR.1 -- [cluster/LIS] -- IPmcR.2 ----- [LAN.2]

   LAN.1 and this
   is reflected by the focus LAN.2 are subnets (such as Ethernet) with attached hosts
   that are members of key examples.  However, group X.

   IPmcR.1 and IPmcR.2 are multicast routers with interfaces to the coding of
   each MARS message means that any higher layer protocol identifiable
   by a two byte Ethernet Type code can LIS.

   A traditional solution would be supported by to treat the LIS as a MARS. unicast subnet,
   and use tunneling routers. However, this would not allow hosts on the
   LIS to participate in the cross-LIS traffic.

   Assume IPmcR.1 is receiving packets promiscuously on its LAN.1
   interface. Assume further it is configured to propagate multicast



Armitage                Expires March 20th, April 24th, 1996                [Page 51] 54]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   As noted in section 4.3, the 16 bit 'Protocol type' (ar$pro) at the
   start of each MARS message is taken from the following two sets:

      0x0000


   traffic to 0x05FF  Reserved all attached interfaces. In this case that means the LIS.

   When a packet for future use by group X arrives on its LAN.1 interface, IPmcR.1
   simply sends the IETF.
      0x0600 packet to 0xFFFF  Protocols defined by the matching Ethertypes.

   Every MARS MUST implement entirely separate logical mapping tables
   and support. Every cluster member must interpret messages from group X on the
   MARS in LIS interface as a normal
   host would (Issuing MARS_REQUEST for group X, creating the context of VC,
   sending the protocol type that packet).

   Assuming IPmcR.2 initialised itself with the MARS message refers
   to.

   Every MARS and MARS client MUST treat Cluster Member IDs in the
   context as a member of the protocol type carried in the MARS message or data
   packet containing the CMI.

   For example, IPv6 has
   entire Class D space, it will have been allocated an Ethertype of 0x86DD.  An IPv6
   multicasting client sets the ar$pro field returned as a member of every MARS message to
   0x86DD.  When carrying IPv6 addresses X
   even if no other nodes on the ar$spln and ar$tpln fields
   are either 0 (for null or non-existent information) or 16 (for the
   full IPv6 address).

   The LLC/SNAP encapsulations described in section 5 similarly allow
   multiple protocols to be identified by the use of different values in
   appropriate encapsulation fields.

   For example, an IPv6 packet is encapsulated as:

      [0xAA-AA-03][0x00-00-5E][0x00-01][pkt$cmi][0x86DD][IPv6 packet]

   A host or endpoint LIS were members. All packets for group
   X received on IPmcR.2's LIS interface that may be retransmitted on LAN.2.

   If IPmcR.1 is using similarly initialised the same MARS to support
   multicasting needs of multiple protocols MUST not assume their CMI reverse process will be the same apply
   for each protocol.

   Values multicast traffic from LAN.2 to LAN.1, for ar$pro in any multicast group.
   The benefit of this scenario is that cluster members within the range 0 to 0x5FF LIS
   may be assigned specific
   interpretations by the IETF also join and leave group X at anytime.

8.4   Joining in future documents.


10.    Supplementary parameter processing.

   MARS messages currently 'semi-promiscuous' mode.

   Both unicast and multicast IP routers have a common header, followed by a variable
   set of fields whose interpretation depends problem -
   limitations on the contents of the
   ar$op field. The format number of these fields, and AAL contexts available at their contents, reflect ATM
   interfaces.  Being 'promiscuous' in the minimum information currently deemed necessary RFC 1112 sense means that for
   every M hosts sending to support the
   MARS model.

   Every MARS message carries an ar$extoff field, immediately following
   the ar$op field. Its role N groups, a multicast router's ATM interface
   will have M*N incoming reassembly engines tied up.

   It is not hard to indicate whether supplementary
   parameters have been supplied along with the basic address mappings.



Armitage                Expires March 20th, 1996                [Page 52]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


   (This mechanism will enable the addition envisage situations where a number of additional functionality
   to multicast
   groups are active within the MARS protocol in later documents.)

   Supplementary parameters LIS but are conveyed as not required to be
   propagated beyond the LIS itself. An example might be a list of TLV (type, length,
   value) encoded information elements.  The TLV(s) begin distributed
   simulation system specifically designed to use the high speed IP/ATM
   environment. There may be no practical way its traffic could be
   utilised on 'the other side' of the first
   32 bit boundary following multicast router, yet under the last 'conventional' field in
   conventional scheme the MARS
   message (e.g. after ar$tsa.N in a MARS_MULTI, after ar$max.N in router would have to be a
   MARS_JOIN, etc).

10.1   Interpreting the ar$extoff field.

   If the bottom 16 bits of leaf to each
   participating host anyway.

   As this problem occurs below the ar$extoff field are non-zero IP layer, it
   indicates is worth noting that
   'scoping' mechanisms at the IP multicast routing level do not provide
   a list of one or more TLVs have been appended solution. An IP level scope would still result in the router's ATM
   interface receiving traffic on the scoped groups, only to drop it.

   In this situation the
   MARS message.  The bottom 16 bits network administrator might configure their
   multicast routers to exclude sections of ar$extoff then represent an
   unsigned integer an offset (in octets) from the beginning of Class D address space
   when issuing MARS_JOIN(s). Multicast groups that will never be
   propagated beyond the MARS
   message (the MSB of cluster will not have the ar$hrd field) to router listed as a
   member, and the first TLV.

   As TLVs are 32 bit aligned router will never have to receive (and simply ignore)
   traffic from those groups.

   Another scenario involves the bottom 2 bits product M*N exceeding the capacity of ar$extoff are a
   single router's interface (especially if the same interface must also
   reserved.



Armitage                Expires April 24th, 1996                [Page 55]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   support a unicast IP router service).

   A receiver MUST mask off these two bits before calculating
   the octet offset network administrator may choose to the TLV list.  A sender MUST set these two bits add a second node, to zero.

   If the bottom 16 bits of ar$extoff are zero no TLVs have been
   appended function
   as a parallel IP multicast router. Each router would be configured to
   be 'promiscuous' over separate parts of the basic MARS message.

   The top 16 bits Class D address space,
   thus exposing themselves to only part of ar$extoff (the two bytes immediately following the
   ar$op field) carry a standard VC load. This sharing
   would be completely transparent to IP checksum calculated across hosts within the
   entire MARS message (which LIS.

   Restricted promiscuous mode does not include the LLC/SNAP header).
   These 16 bits are set to zero before performing break RFC 1112's use of IGMP
   Report messages. If the calculation prior router is configured to transmission of serve a message.

   As given block
   of Class D addresses, it will receive the entire LLC/SNAP encapsulated MARS message IGMP Report.  If the router
   is protected by not configured to support a given block, then the
   32 bit CRC existence of the AAL5 transport, implementors MAY choose an
   IGMP Report for a group in that block is irrelevant to ignore the checksum facility. If no checksum is calculated these bits MUST
   be reset before transmission.

   Implementations that do not implement checksumming MUST silently
   discard messages received with these bits non-zero.  Otherwise, if
   these bits router.
   All routers are non-zero on reception able to track membership changes through the receiver MUST perform
   MARS_JOIN and MARS_LEAVE traffic anyway. (Section 8.5 discusses a
   checksum computation
   better alternative to IGMP within a cluster.)

   Mechanisms and discard the packet silently if the checksums
   do not match. The checksum bits are treated as zero reasons for the purpose establishing these modes of the receive side checksum calculation.

   Checksum generation and processing operation are NOT REQUIRED for conformance
   with
   beyond the scope of this document.






Armitage                Expires March 20th, 1996                [Page 53]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


10.2   The format

8.5   An alternative to IGMP Queries.

   An unfortunate aspect of TLVs.

   When they exist, TLVs begin on 32 bit boundaries, are multiples IGMP is that it assumes multicasting of 32
   bits in length, IP
   packets is a cheap and form trivial event at the link layer. As a sequential list terminated
   consequence, regular IGMP Queries are multicasted by a NULL TLV.

   The TLV structure is:

      [Type - 2 octets][Length - 2 octets][Value - n*4 octets]

   The Type subfield indicates how the contents of the Value subfield routers to group
   224.0.0.1. These queries are intended to be interpreted.

   The Length subfield indicates the number trigger IGMP Replies by
   cluster members that have layer 3 members of VALID octets in the Value
   subfield. Valid octets in the Value subfield start immediately after
   the Length subfield. particular groups.

   The offset (in octets) from the start of this
   TLV MARS_GROUPLIST_REQUEST and MARS_GROUPLIST_REPLY messages were
   designed to allow routers to avoid actually transmitting IGMP Queries
   out into a cluster.

   Whenever the start of the next TLV router's forwarding engine wishes to transmit an IGMP
   query, a MARS_GROUPLIST_REQUEST can be sent to the MARS instead. The
   resulting MARS_GROUPLIST_REPLY(s) (described in section 5.3) from the list is given by
   MARS carry all the
   following formula:

      offset = (length + 4 + ((4-(length & 3)) % 4))

   (where % is information that the modulus operator)

   The Value subfield router would have ascertained
   from IGMP replies.

   It is padded with 0, 1, 2, or 3 octets RECOMMENDED that multicast routers utilise this MARS service to ensure
   minimise IGMP traffic within the
   next TLV is 32 bit aligned. The padded locations MUST be set to zero.

   (For example, cluster.

   By default a TLV MARS_GROUPLIST_REQUEST SHOULD specify the entire address
   space (e.g. <224.0.0.0, 239.255.255.255> in an IPv4 environment).
   However, routers serving part of the address space (as described in
   section 8.4) MAY choose to issue MARS_GROUPLIST_REQUESTs that needed specify
   only 5 valid octets the subset of information
   would be 12 octets long. The Length subfield would hold the value 5,
   and address space they are serving.

   (On the Value subfield surface it would be padded out also seem useful for multicast routers to 8 bytes.  The 5 valid
   octets of information begin at the first octet



Armitage                Expires April 24th, 1996                [Page 56]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   track MARS_JOINs and MARS_LEAVEs that arrive with ar$flags.layer3grp
   set. These might be used in lieu of IGMP Reports, to provide the Value
   subfield.)

   The Type subfield is formatted in
   router with timely indication that a new layer 3 group member exists
   within the following way:

          |   1st octet   |   2nd octet   |
           7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0
          +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
          | x |               y           |
          +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   The most significant 2 bits (Type.x) determine how cluster. However, this only works on VC mesh supported
   groups, and is therefore NOT recommended).

   Appendix B discusses less elegant mechanisms for reducing the impact
   of IGMP traffic within a recipient should
   behave when it doesn't recognise cluster, on the TLV type indicated assumption that the IP/ATM
   interfaces to the cluster are being used by un-optimised IP
   multicasting code.

8.6   CMIs across multiple interfaces.

   The Cluster Member ID is only unique within the Cluster managed by a
   given MARS. On the lower
   14 bits (Type.y). surface this might appear to leave us with a
   problem when a multicast router is routing between two or more
   Clusters using a single physical ATM interface.  The required behaviours are:

      Type.x = 0   Skip router will
   register with two or more MARSs, and thereby acquire two or more
   independent CMI's. Given that each MARS has no reason to synchronise
   their CMI allocations, it is possible for a host in one cluster to
   have the TLV, continue processing same CMI has the list.
      Type.x = 1   Stop processing, silently drop router's interface to another Cluster. How
   does the MARS message.
      Type.x = 2   Stop processing, drop message, give error indication.
      Type.x = 3   Reserved. (currently treat as x = 0) router distinguish between its own reflected packets, and
   packets from that other host?

   The answer lies in the fact that routers (and hosts) actually
   implement logical IP/ATM interfaces over a single physical ATM
   interface. Each logical interface will have a unique ATM Address (eg.
   an NSAP with different SELector fields, one for each logical
   interface).

   Each logical IP/ATM interface is configured with the address of a
   single MARS, attaches to only one cluster, and so had only one CMI to
   worry about. Each of the MARSs that the router is registered with
   will have been given a different ATM Address (corresponding to the
   different logical IP/ATM interfaces) in each registration MARS_JOIN.

   When hosts in a cluster add the router as a leaf node, they'll
   specify the ATM Address of the appropriate logical IP/ATM interface
   on the router in the L_MULTI_ADD message. Thus, each logical IP/ATM
   interface will only have to check and filter on CMIs assigned by its
   own MARS.

   In essence the cluster differentiation is achieved by ensuring that
   logical IP/ATM interfaces are assigned different ATM Addresses.







Armitage                Expires April 24th, 1996                [Page 57]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


9.    Multiprotocol applications of the MARS and MARS clients.

   A deliberate attempt has been made to describe the MARS and
   associated mechanisms in a manner independent of a specific higher
   layer protocol being run over the ATM cloud. The immediate
   application of this document will be in an IPv4 environment, and this
   is reflected by the focus of key examples.  However, the ar$pro.type
   and ar$pro.snap fields in every MARS control message allow any higher
   layer protocol that has a 'short form' or 'long form' of protocol
   identification (section 4.3) to be supported by a MARS.

   Every MARS MUST implement entirely separate logical mapping tables
   and support. Every cluster member must interpret messages from the
   MARS in the context of the protocol type that the MARS message refers
   to.

   Every MARS and MARS client MUST treat Cluster Member IDs in the
   context of the protocol type carried in the MARS message or data
   packet containing the CMI.

   For example, IPv6 has been allocated an Ethertype of 0x86DD.  This
   means the 'short form' of protocol identification must be used in the
   MARS control messages and the data path encapsulation (section 5.5).
   An IPv6 multicasting client sets the ar$pro.type field of every MARS
   message to 0x86DD.  When carrying IPv6 addresses the ar$spln and
   ar$tpln fields are either 0 (for null or non-existent information) or
   16 (for the full IPv6 address).

   Following the rules in section 5.5, an IPv6 data packet is
   encapsulated as:

      [0xAA-AA-03][0x00-00-5E][0x00-01][pkt$cmi][0x86DD][IPv6 packet]

   A host or endpoint interface that is using the same MARS to support
   multicasting needs of multiple protocols MUST not assume their CMI
   will be the same for each protocol.


10.    Supplementary parameter processing.

   The ar$extoff field in the [Fixed header] indicates whether
   supplementary parameters are being carried by a MARS control message.
   This mechanism is intended to enable the addition of new
   functionality to the MARS protocol in later documents.

   Supplementary parameters are conveyed as a list of TLV (type, length,
   value) encoded information elements.  The TLV(s) begin on the first
   32 bit boundary following the [Addresses] field in the MARS control



Armitage                Expires April 24th, 1996                [Page 58]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   message (e.g. after ar$tsa.N in a MARS_MULTI, after ar$max.N in a
   MARS_JOIN, etc).

10.1   Interpreting the ar$extoff field.

   If the ar$extoff field is non-zero it indicates that a list of one or
   more TLVs have been appended to the MARS message.  The first TLV is
   found by treating ar$extoff as an unsigned integer representing an
   offset (in octets) from the beginning of the MARS message (the MSB of
   the ar$hrd field).

   As TLVs are 32 bit aligned the bottom 2 bits of ar$extoff are also
   reserved. A receiver MUST mask off these two bits before calculating
   the octet offset to the TLV list.  A sender MUST set these two bits
   to zero.

   If ar$extoff is zero no TLVs have been appended.

10.2   The format of TLVs.

   When they exist, TLVs begin on 32 bit boundaries, are multiples of 32
   bits in length, and form a sequential list terminated by a NULL TLV.

   The TLV structure is:

      [Type - 2 octets][Length - 2 octets][Value - n*4 octets]

   The Type subfield indicates how the contents of the Value subfield
   are to be interpreted.

   The Length subfield indicates the number of VALID octets in the Value
   subfield. Valid octets in the Value subfield start immediately after
   the Length subfield.  The offset (in octets) from the start of this
   TLV to the start of the next TLV in the list is given by the
   following formula:

      offset = (length + 4 + ((4-(length & 3)) % 4))

   (where % is the modulus operator)

   The Value subfield is padded with 0, 1, 2, or 3 octets to ensure the
   next TLV is 32 bit aligned. The padded locations MUST be set to zero.

   (For example, a TLV that needed only 5 valid octets of information
   would be 12 octets long. The Length subfield would hold the value 5,
   and the Value subfield would be padded out to 8 bytes.  The 5 valid
   octets of information begin at the first octet of the Value
   subfield.)



Armitage                Expires April 24th, 1996                [Page 59]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   The Type subfield is formatted in the following way:

          |   1st octet   |   2nd octet   |
           7 6 5 4 3 2 1 0 7 6 5 4 3 2 1 0
          +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
          | x |               y           |
          +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

   The most significant 2 bits (Type.x) determine how a recipient should
   behave when it doesn't recognise the TLV type indicated by the lower
   14 bits (Type.y). The required behaviours are:

      Type.x = 0   Skip the TLV, continue processing the list.
      Type.x = 1   Stop processing, silently drop the MARS message.
      Type.x = 2   Stop processing, drop message, give error indication.
      Type.x = 3   Reserved. (currently treat as x = 0)

   (The error indication generated when Type.x = 2 SHOULD be logged in
   some locally significant fashion. Consequential MARS message activity
   in response to such an error condition will be defined in future
   documents.)

   The TLV type space (Type.y) is further subdivided to encourage use
   outside the IETF.

      0                       Null TLV.
      0x0001 - 0x0FFF         Reserved for the IETF.
      0x1000 - 0x11FF         Allocated to the ATM Forum.
      0x1200 - 0x37FF         Reserved for the IETF.
      0x3800 - 0x3FFF         Experimental use.

10.3   Processing MARS messages with TLVs.

   Supplementary parameters act as modifiers to the basic behaviour
   specified by the ar$op field of any given MARS message.

   If a MARS message arrives with a non-zero ar$extoff field its TLV
   list MUST be parsed before handling the MARS message in accordance
   with the ar$op value. Unrecognised TLVs MUST be handled as required
   by their Type.x value.

   How TLVs modify basic MARS operations will be ar$op and TLV specific.

10.4   Initial set of TLV elements.

   Conformance with this document only REQUIRES the recognition of one
   TLV, the Null TLV. This terminates a list of TLVs, and MUST be
   present if ar$extoff is non-zero in a MARS message. It MAY be the



Armitage                Expires April 24th, 1996                [Page 60]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   only TLV present.

   The Null TLV is coded as:

      [0x00-00][0x00-00]

   Future documents will describe the formats, contents, and
   interpretations of additional TLVs. The minimal parsing requirements
   imposed by this document are intended to allow conformant MARS and
   MARS client implementations to deal gracefully and predictably with
   future TLV developments.

11.    Key Decisions and open issues.

   The key decisions this document proposes:

      A Multicast Address Resolution Server (MARS) is proposed to co-
      ordinate and distribute mappings of ATM endpoint addresses to
      arbitrary higher layer 'multicast group addresses'. The specific
      case of IPv4 multicast is used as the example.

      The concept of 'clusters' is introduced to define the scope of a
      MARS's responsibility, and the set of ATM endpoints willing to
      participate in link level multicasting.

      A MARS is described with the functionality required to support
      intra-cluster multicasting using either VC meshes or ATM level
      multicast servers (MCSs).

      LLC/SNAP encapsulation of MARS control messages allows MARS and
      ATMARP traffic to share VCs, and allows partially co-resident MARS
      and ATMARP entities.

      New message types:

         MARS_JOIN, MARS_LEAVE, MARS_REQUEST. Allow endpoints to join,
         leave, and request the current membership list of multicast
         groups.

         MARS_MULTI. Allows multiple ATM addresses to be returned by the
         MARS in response to a MARS_REQUEST.

         MARS_MSERV, MARS_UNSERV. Allow multicast servers to register
         and deregister themselves with the MARS.

         MARS_SJOIN, MARS_SLEAVE. Allow MARS to pass on group membership
         changes to multicast servers.




Armitage                Expires April 24th, 1996                [Page 61]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


         MARS_GROUPLIST_REQUEST, MARS_GROUPLIST_REPLY.  Allow MARS to
         indicate which groups have actual layer 3 members. May be used
         to support IGMP in IPv4 environments, and similar functions in
         other environments.

         MARS_REDIRECT_MAP.  Allow MARS to specify a set of backup MARS
         addresses.

      'wild card' MARS mapping table entries possible, where a single
      ATM address is simultaneously associated with blocks of multicast
      group addresses.

   For the MARS protocol ar$op.version = 0. The complete set of MARS
   control messages and ar$op.type values is:

      1   MARS_REQUEST
      2   MARS_MULTI
      3   MARS_MSERV
      4   MARS_JOIN
      5   MARS_LEAVE
      6   MARS_NAK
      7   MARS_UNSERV
      8   MARS_SJOIN
      9   MARS_SLEAVE
      10  MARS_GROUPLIST_REQUEST
      11  MARS_GROUPLIST_REPLY
      12  MARS_REDIRECT_MAP

   A number of issues are left open at this stage, and are likely to be
   the subject of on-going research and additional documents that build
   upon this one.

      The specified endpoint behaviour allows the use of
      redundant/backup MARSs within a cluster. However, no
      specifications yet exist on how these MARSs co-ordinate amongst
      themselves. (The default is to only have one MARS per cluster.)

      The specified endpoint behaviour and MARS service allows the use
      of multiple MCSs per group.  However, no specifications yet exist
      on how this may be used, or how these MCSs co-ordinate amongst
      themselves.  Until futher work is done on MCS co-ordination
      protocols the default is to only have one MCS per group.

      The MARS relies on the cluster member dropping off
      ClusterControlVC if the cluster member dies. It is not clear if
      additional mechanisms are needed to detect and delete 'dead'
      cluster members.




Armitage                Expires April 24th, 1996                [Page 62]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


      If a multicast server attempts to MARS_MSERV for an existing VC
      mesh supported group, it would be nice to have current senders to
      the group migrate their outgoing VCs from the actual cluster
      member leaf nodes to the newly registered multicast server(s). How
      this might be achieved, the load this would place on the MARS, and
      its scalability, is for further study.

      Supporting layer 3 'broadcast' as a special case of multicasting
      (where the 'group' encompasses all cluster members) has not been
      explicitly discussed.

      Supporting layer 3 'unicast' as a special case of multicasting
      (where the 'group' is a single cluster member, identified by the
      cluster member's unicast protocol address) has not been explicitly
      discussed.

      The future development of ATM Group Addresses and Leaf Initiated
      Join to ATM Forum's UNI specification has not been addressed.
      (However, the problems identified in this document with respect to
      VC scarcity and impact on AAL contexts will not be fixed by such
      developments in the signalling protocol.)


Security Consideration

   Security consideration are not addressed in this document.

Acknowledgments

   The discussions within the IP over ATM Working Group have helped
   clarify the ideas expressed in this document. John Moy (Cascade
   Communications Corp.) initially suggested the idea of wild-card
   entries in the ARP Server.  Drew Perkins (Fore Systems) provided
   rigorous and useful critique of early proposed mechanisms for
   distributing and validating group membership information.  Susan
   Symington (and co-workers at MITRE Corp., Don Chirieleison, Rich
   Verjinski, and Bill Barns) clearly articulated the need for multicast
   server support, proposed a solution, and challenged earlier block
   join/leave mechanisms. John Shirron (Fore Systems) provided useful
   improvements on my original revalidation procedures.

   Susan Symington and Bryan Gleeson (Adaptec) independently championed
   the need for the service provided by MARS_GROUPLIST_REQUEST/REPLY.
   The new encapsulation scheme arose from WG discussions, captured by
   Bryan Gleeson in an interim Internet Draft (with Keith McCloghrie
   (Cisco), Andy Malis (Ascom Nexion), and Andrew Smith (Bay Networks)
   as key contributors).  James Watt (Newbridge) and Joel Halpern
   (Newbridge) motivated the development of a more multiprotocol MARS



Armitage                Expires April 24th, 1996                [Page 63]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   control message format, evolving it away from its original ATMARP
   roots.  They also motivated the development of Type #1 and Type #2
   data path encapsulations.

   Finally, Jim Rubas (IBM) supplied the MARS pseudo-code in Appendix F.


Author's Address

   Grenville Armitage
   Bellcore, 445 South Street
   Morristown, NJ, 07960
   USA

   Email: gja@thumper.bellcore.com
   Ph. +1 201 829 2635


References
   [1] S. Deering, "Host Extensions for IP Multicasting", RFC 1112,
   Stanford University, August 1989.

   [2] Heinanen, J., "Multiprotocol Encapsulation over ATM Adaption
   Layer 5", RFC 1483, USC/Information Science Institute, July 1993.

   [3] Laubach, M., "Classical IP and ARP over ATM", RFC1577, Hewlett-
   Packard Laboratories, December 1993

   [4] ATM Forum, "ATM User Network Interface (UNI) Specification
   Version 3.1", ISBN 0-13-393828-X, Prentice Hall, Englewood Cliffs,
   NJ, June 1995.

   [5] D. Waitzman, C. Partridge, S. Deering, "Distance Vector Multicast
   Routing Protocol", RFC 1075, November 1988.

   [6] M. Perez, F. Liaw, D. Grossman, A. Mankin, E. Hoffman, A. Malis,
   "ATM Signaling Support for IP over ATM", RFC 1755, February 1995.

   [7] M. Borden, E. Crawley, B. Davie, S. Batsell, "Integration of
   Real-time Services in an IP-ATM Network Architecture.", RFC 1821,
   August 1995.

   [8] ATM Forum, "ATM User-Network Interface Specification Version
   3.0", Englewood Cliffs, NJ: Prentice Hall, September 1993







Armitage                Expires March 20th, April 24th, 1996                [Page 54] 64]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   (The error indication generated when Type.x = 2 SHOULD be logged in
   some locally significant fashion. Consequential MARS message activity
   in response to such an error condition will be defined in future
   documents.)

   The TLV type space (Type.y) is further subdivided to encourage use
   outside the IETF.

      0                       Null TLV.
      0x0001 - 0x0FFF         Reserved for the IETF.
      0x1000 - 0x11FF         Allocated


Appendix A.  Hole punching algorithms.

   Implementations are entirely free to the ATM Forum.
      0x1200 - 0x37FF         Reserved for the IETF.
      0x3800 - 0x3FFF         Experimental use.

10.3   Processing MARS messages comply with TLVs.

   Supplementary parameters act as modifiers to the basic behaviour
   specified by the ar$op field body of this
   memo in any given MARS message.

   If a way they see fit. This appendix is purely for
   clarification.

   A MARS message arrives with implementation might pre-construct a non-zero ar$extoff field its TLV
   list MUST be parsed before handling the MARS message in accordance
   with the ar$op value. Unrecognised TLVs must be handled as required
   by their Type.x value.

   How TLVs modify basic MARS operations will be ar$op and TLV specific.

10.4   Initial set of TLV elements.

   Conformance with this document only REQUIRES <min,max> pairs
   (P) that reflects the recognition entire Class D space, excluding any addresses
   currently supported by multicast servers. The <min> field of one
   TLV, the Null TLV. This terminates a list
   first pair MUST be 224.0.0.0, and the <max> field of TLVs, the last pair
   must be 239.255.255.255. The first and MUST last pair may be
   present if ar$extoff the same.
   This set is non-zero in updated whenever a multicast server registers or
   deregisters.

   When the MARS message. It MAY be must perform 'hole punching' it might consider the
   only TLV present.

   It is coded simply as:

      [0x00-00][0x00-00]

   Future documents will describe
   following algorithm:

      Assume the formats, contents, and
   interpretations of additional TLVs. The minimal parsing requirements
   imposed MARS_JOIN/LEAVE received by this document are intended to allow conformant the MARS from the cluster
      member specied the block <Emin, Emax>.

      Assume Pmin(N) and
   MARS client implementations to deal gracefully Pmax(N) are the <min> and predictably with
   future TLV developments.

11.    Key Decisions <max> fields from the
      Nth pair in the MARS's current set P.

      Assume set P has K pairs. Pmin(1) MUST equal 224.0.0.0, and open issues.

   The key decisions this document proposes:

      A Multicast Address Resolution Server (MARS)
      Pmax(M) MUST equal 239.255.255.255. (If K == 1 then no hole
      punching is proposed to co-



Armitage                Expires March 20th, 1996                [Page 55]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


      ordinate and distribute mappings of ATM endpoint addresses to
      arbitrary higher layer 'multicast group addresses'. The specific
      case required).

      Execute pseudo-code:

         create copy of IPv4 multicast set P, call it set C.

         index1 = 1;
         while (Pmax(index1) <= Emin)
            index1++;

         index2 = K;
         while (Pmin(index2) >= Emax)
            index2--;

         if (index1 > index2)
            Exit, as the hole-punched set is null.

         if (Pmin(index1) < Emin)
            Cmin(index1) = Emin;

         if (Pmax(index2) > Emax)
            Cmax(index2) = Emax;




Armitage                Expires April 24th, 1996                [Page 65]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


         Set C is used as the example.

      The concept required 'hole punched' set of 'clusters' is introduced to define address blocks.

   The resulting set C retains all the scope of a MARS's responsibility, and the set of ATM endpoints willing to
      participate in link level multicasting.

      A MARS is described with pre-constructed 'holes'
   covering the functionality required to support
      intra-cluster multicasting using either VC meshes or ATM level multicast servers (MCSs).

      MARS message formats and encapsulation allow co-resident MARS and
      ATM ARP Server implementations.

      New message types:

         MARS_JOIN, MARS_LEAVE, MARS_REQUEST. Allow endpoints servers, but will have been pruned to join,
         leave, and request cover
   the current membership list section of multicast
         groups.

         MARS_MULTI. Allows multiple ATM addresses to be returned the Class D space specified by the
         MARS in response to originating host's
   <Emin,Emax> values.

   The host end should keep a MARS_REQUEST.

         MARS_MSERV, MARS_UNSERV. Allow multicast servers to register
         and deregister themselves with the MARS.

         MARS_SJOIN, MARS_SLEAVE. Allow MARS to pass on group membership
         changes to multicast servers.

         MARS_GROUPLIST_REQUEST, MARS_GROUPLIST_REPLY.  Allow MARS to
         indicate which groups have actual layer 3 members. May be used
         to support IGMP in IPv4 environments, and similar functions table, H, of open VCs in
         other environments.

         MARS_REDIRECT_MAP.  Allow MARS to specify a set ascending order
   of backup MARS
         addresses.

      'wild card' MARS mapping table entries possible, where a single
      ATM address Class D address.

      Assume H(x).addr is simultaneously the Class address associated with blocks of multicast
      group addresses. VC.x.
      Assume H(x).addr < H(x+1).addr.

   The complete set pseudo code for updating VCs based on an incoming JOIN/LEAVE
   might be:

      x = 1;
      N = 1;

      while (x < no.of VCs open)
      {
            while (H(x).addr > max(N))
            {
                  N++;
                  if (N > no. of messages, and ar$op values, is:

      11   MARS_REQUEST
      12   MARS_MULTI
      13   MARS_MSERV
      14   MARS_JOIN pairs in JOIN/LEAVE)
                        return(0);
            }

            if ((H(x).addr <= max(N) &&
                        ((H(x).addr >= min(N))
                              perform_VC_update();
            x++;
      }


















Armitage                Expires March 20th, April 24th, 1996                [Page 56] 66]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


      15   MARS_LEAVE
      16   MARS_NAK
      17   MARS_UNSERV
      18   MARS_SJOIN
      19   MARS_SLEAVE
      20   MARS_GROUPLIST_REQUEST
      21   MARS_GROUPLIST_REPLY
      22   MARS_REDIRECT_MAP

   A number


Appendix B.  Minimising the impact of IGMP in IPv4 environments.

   Implementing any part of issues are left open at this stage, and are likely appendix is not required for
   conformance with this document.  It is provided solely to be
   the subject of on-going research and additional documents document
   issues that build
   upon this one. have been identified.

   The specified endpoint behaviour allows the use intent of
      redundant/backup MARSs within a cluster. However, no
      specifications yet exist on how these MARSs co-ordinate amongst
      themselves. (The default section 5.1 is for cluster members to only have one MARS per cluster.)
   outgoing point to multipoint VCs when they are actually sending data
   to a particular multicast groups. However, in most IPv4 environments
   the multicast routers attached to a cluster will periodically issue
   IGMP Queries to ascertain if particular groups have members.  The specified endpoint behaviour
   current IGMP specification attempts to avoid having every group
   member respond by insisting that each group member wait a random
   period, and MARS service allows responding if no other member has responded before them.
   The IGMP reply is sent to the use multicast address of multiple MCSs per the group being
   queried.

   Unfortunately, as it stands the IGMP algorithm will be a nuisance for
   cluster members that are essentially passive receivers within a given
   multicast group.  However, It is just as likely that a passive member, with no specifications yet exist
      on how this may
   outgoing VC already established to the group, will decide to send an
   IGMP reply - causing a VC to be used, or how these MCSs co-ordinate amongst
      themselves.  Until futher work established were there was no need
   for one. This is done not a fatal problem for small clusters, but will
   seriously impact on MCS co-ordination
      protocols the default is ability of a cluster to only have one MCS per group. scale.

   The MARS relies on the cluster member dropping off
      ClusterControlVC if the cluster member dies. It most obvious solution is not clear if
      additional mechanisms are needed for routers to detect use the
   MARS_GROUPLIST_REQUEST and delete 'dead' MARS_GROUPLIST_REPLY messages, as
   described in section 8.5. This would remove the regular IGMP Queries,
   resulting in cluster members.

      If members only sending an IGMP Report when they
   first join a multicast server attempts group.

   Alternative solutions do exist. One would be to MARS_MSERV modify the IGMP reply
   algorithm, for an existing example:

      If the group member has VC
      mesh supported group, it would be nice open to the group proceed as per RFC
      1112 (picking a random reply delay between 0 and 10 seconds).

      If the group member does not have current senders VC already open to the group,
      pick random reply delay between 10 and 20 seconds instead, and
      then proceed as per RFC 1112.

   If even one group migrate their outgoing VCs from the actual cluster member leaf nodes is sending to the newly registered multicast server(s). How
      this might be achieved, the load this would place on group at the MARS, and
      its scalability, have not yet been considered.

      Supporting layer 3 'broadcast' as a special case of multicasting
      (where time the 'group' encompasses IGMP
   Query is issued then all cluster members) has not been
      explicitly discussed.

      Supporting layer 3 'unicast' as a special case of multicasting
      (where the 'group' is a single cluster member, identified by passive receivers will find the
      cluster member's unicast protocol address) IGMP
   Reply has not been explicitly
      discussed.

      The future development transmitted before their delay expires, so no new VC
   is required. If all group members are passive at the time of ATM Group Addresses and Leaf Initiated
      Join to ATM Forum's UNI specification has not been addressed.
      (However, the problems identified in this document with respect to IGMP
   Query then a response will eventually arrive, but 10 seconds later
   than under conventional circumstances.




Armitage                Expires March 20th, April 24th, 1996                [Page 57] 67]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   The preceeding solution requires re-writing existing IGMP code, and
   implies the ability of the IGMP entity to ascertain the status of VCs
   on the underlying ATM interface. This is not likely to be available
   in the short term.

   One short term solution is to provide something like the preceeding
   functionality with a 'hack' at the IP/ATM driver level within cluster
   members. Arrange for the IP/ATM driver to snoop inside IP packets
   looking for IGMP traffic. If an IGMP packet is accepted for
   transmission, the IP/ATM driver can buffer it locally if there is no
   VC scarcity already active to that group. A 10 second timer is started, and impact if
   an IGMP Reply for that group is received from elsewhere on AAL contexts will not be fixed by such
      developments in the signalling protocol.)


Security Consideration

   Security consideration are not addressed in this document.

Acknowledgments

   The discussions within
   cluster the IP over ATM Working Group have helped
   clarify timer is reset. If the ideas expressed in this document. John Moy (Cascade
   Communications Corp.) initially suggested timer expires, the idea of wild-card
   entries in IP/ATM driver
   then establishes a VC to the ARP Server.  Drew Perkins (Fore Systems) provided
   rigorous and useful critique of early proposed mechanisms group as it would for
   distributing and validating a normal IP
   multicast packet.

   Some network implementors may find it advantageous to configure a
   multicast server to support the group membership information.  Susan
   Symington (and co-workers at MITRE Corp., Don Chirieleison, Rich
   Verjinski, and Bill Barns) clearly articulated 224.0.0.1, rather than rely on
   a mesh. Given that IP multicast routers regularly send IGMP queries
   to this address, a mesh will mean that each router will permanently
   consume an AAL context within each cluster member. In clusters served
   by multiple routers the need for VC load within switches in the underlying ATM
   network will become a scaling problem.

   Finally, if a multicast server support, proposed is used to support 224.0.0.1, another
   ATM driver level hack becomes a solution, and challenged earlier block
   join/leave mechanisms. John Shirron (Fore Systems) provided useful
   improvements on my original revalidation procedures. Susan Symington
   and Bryan Gleeson (Adaptec) independently championed possible solution to IGMP Reply
   traffic.  The ATM driver may choose to grab all outgoing IGMP packets
   and send them out on the need VC established for sending to 224.0.0.1,
   regardless of the
   service provided by MARS_GROUPLIST_REQUEST/REPLY.  The new
   encapsulation scheme arose from WG discussions, captured by Bryan
   Gleeson in an interim Internet Draft (with Keith McCloghrie (Cisco),
   Andy Malis (Ascom Nexion), and Andrew Smith (Bay Networks) as key
   contributors).

Author's Address

   Grenville Armitage
   MRE 2P255, 445 South Street
   Morristown, NJ, 07960
   USA

   Email: gja@thumper.bellcore.com
   Ph. +1 201 829 2635


References
   [1] S. Deering, "Host Extensions for IP Multicasting", RFC 1112,
   Standford University, August 1989.

   [2] Heinanen, J., "Multiprotocol Encapsulation over ATM Adaption
   Layer 5", RFC 1483, USC/Information Science Institute, July 1993.

   [3] Laubach, M., "Classical IP Class D address the IGMP message was actually for.
   Given that all hosts and ARP over ATM", RFC1577, Hewlett-
   Packard Laboratories, December 1993 routers must be members of 224.0.0.1, the
   intended recipients will still receive the IGMP Replies. The negative
   impact is that all cluster members will receive the IGMP Replies.



















Armitage                Expires March 20th, April 24th, 1996                [Page 58] 68]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


   [4] ATM Forum, "ATM User Network Interface (UNI) Specification
   Version 3.1", ISBN 0-13-393828-X, Prentice Hall, Englewood Cliffs,
   NJ, June 1995.

   [5] D. Waitzman,


Appendix C. Partridge, S. Deering, "Distance Vector Multicast
   Routing Protocol", RFC 1075, November 1988.

   [6] M. Perez, F. Liaw, D. Grossman, A. Mankin, E. Hoffman, A. Malis,
   "ATM Signaling Support   Further comments on 'Clusters'.

   The cluster concept was introduced in section 1 for two reasons.  The
   more well known term of Logical IP over ATM", RFC 1755, February 1995.

   [7] M. Borden, E. Crawley, B. Davie, S. Batsell, "Integration Subnet is both very IP specific,
   and constrained to unicast routing boundaries. As the architecture
   described in this document may be re-used in non-IP environments a
   more neutral term was needed. As the needs of
   Real-time Services multicasting are not
   always bound by the same scopes as unicasting, it was not immediately
   obvious that apriori limiting ourselves to LISs was beneficial in the
   long term.

   It must be stressed that Clusters are purely an IP-ATM Network Architecture.", RFC 1821,
   August 1995.






































Armitage                Expires March 20th, 1996                [Page 59]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


Appendix A.  Hole punching algorithms.

   Implementations administrative being.
   You choose their size (i.e. the number of endpoints that register
   with the same MARS) based on your multicasting needs, and the
   resource consumption you are entirely free willing to comply with put up with. The larger the body
   number of this
   memo in any way they see fit. This appendix is purely for
   clarification.

   A MARS implementation ATM attached hosts you require multicast support for, the
   more individual clusters you might pre-construct a set of <min,max> pairs
   (P) choose to establish (along with
   multicast routers to provide inter-cluster traffic paths).

   Given that reflects not all the entire Class D space, excluding hosts in any addresses
   currently supported by given LIS may require multicast servers. The <min> field of
   support, it becomes conceivable that you might assign a single MARS
   to support hosts from across multiple LISs. In effect you have a
   cluster covering multiple LISs, and have achieved 'cut through'
   routing for multicast traffic. Under these circumstances increasing
   the
   first pair MUST geographical size of a cluster might be 224.0.0.0, and considered a good thing.

   However, practical considerations limit the <max> field size of the last pair
   must be 239.255.255.255. The first and last pair clusters.  Having
   a cluster span multiple LISs may not always be the same.
   This set is updated whenever a multicast server registers or
   deregisters.

   When particular 'win'
   situation.  As the MARS must perform 'hole punching' number of multicast capable hosts in your LISs
   increases it might consider becomes more likely that you'll want to constrain a
   cluster's size and force multicast traffic to aggregate at multicast
   routers scattered across your ATM cloud.

   Finally, multi-LIS clusters require a degree of care when deploying
   IP multicast routers. Under the
   following algorithm:

      Assume Classical IP model you need unicast
   routers on the MARS_JOIN/LEAVE received by edges of LISs. Under the MARS from architecture you only
   need multicast routers at the edges of clusters. If your cluster
      member specied
   spans multiple LISs, then the block <Emin, Emax>.

      Assume Pmin(N) and Pmax(N) are multicast routers will perceive
   themselves to have a single interface that is simultaneously attached
   to multiple unicast subnets. Whether this situation will work depends
   on the <min> inter-domain multicast routing protocols you use, and <max> fields from the
      Nth pair in your
   multicast router's ability to understand the MARS's current set P.

      Assume set P has K pairs. Pmin(1) MUST equal 224.0.0.0, new relationship between
   unicast and
      Pmax(M) MUST equal 239.255.255.255. (If K == 1 then no hole
      punching is required).

      Execute pseudo-code:

         create copy multicast topologies.

   In the absence of set P, call it set C.

         index1 = 1;
         while (Pmax(index1) <= Emin)
            index1++;

         index2 = K;
         while (Pmin(index2) >= Emax)
            index2--;

         if (index1 > index2)
            Exit, futher research in this area, networks deployed in
   conformance to this document MUST make their IP cluster and IP LIS
   coincide, so as the hole-punched set is null.

         if (Pmin(index1) < Emin)
            Cmin(index1) = Emin;

         if (Pmax(index2) > Emax)
            Cmax(index2) = Emax; to avoid these complications.




Armitage                Expires March 20th, April 24th, 1996                [Page 60] 69]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


         Set C is the required 'hole punched' set of address blocks.


Appendix D.  TLV list parsing algorithm.

   The resulting set C retains all the MARS's pre-constructed 'holes'
   covering the multicast servers, but will have been pruned to cover following pseudo-code represents how the TLV list format
   described in section of the Class D space specified 10 could be handled by the originating host's
   <Emin,Emax> values.

   The host end should keep a table, H, of open VCs in ascending order
   of Class D address.

      Assume H(x).addr is the Class address associated with VC.x.
      Assume H(x).addr < H(x+1).addr.

   The pseudo code for updating VCs based on an incoming JOIN/LEAVE
   might be:

      x MARS or MARS client.

      list = 1;
      N (ar$extoff & 0xFFFC);

      if (list == 0) exit;

      list = 1; list + message_base;

      while (x < no.of VCs open) (list->Type.y != 0)
            {
            while (H(x).addr > max(N))
                  switch (list->Type.y)
                  {
                        default:
                          {
                  N++;
                           if (N > no. of pairs in JOIN/LEAVE)
                        return(0);
            } (list->Type.x == 0) break;

                           if ((H(x).addr <= max(N) &&
                        ((H(x).addr >= min(N))
                              perform_VC_update();
            x++; (list->Type.x == 1) exit;

                           if (list->Type.x == 2) log-error-and-exit;
                          }

                        [...other handling goes here..]

                  }

                  list += (list->Length + 4 + ((4-(list->Length & 3)) %
                  4));

            }

      return;

















Armitage                Expires March 20th, April 24th, 1996                [Page 61] 70]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


Appendix B.  Minimising the impact E.  Summary of IGMP in IPv4 environments.

   Implementing any part timer values.

   This appendix summarises of various timers or limits mentioned in the
   main body of this appendix is not required for
   conformance with this the document.  It is provided solely to document
   issues that have been identified.

   The intent of section 5.1 is for cluster members to only have
   outgoing point to multipoint VCs when they Values are actually sending data
   to a particular multicast groups. However, specified in most IPv4 environments the multicast routers attached to following
   format:  [x, y, z] indicating a cluster will periodically issue
   IGMP Queries to ascertain if particular groups have members.  The
   current IGMP specification attempts to avoid having every group
   member respond by insisting that each group member wait minimum value of x, a random
   period, recommended
   value of y, and responding if no other member has responded before them.
   The IGMP reply is sent to the multicast address a maximum value of the group being
   queried.

   Unfortunately, as it stands the IGMP algorithm z. A '-' will be a nuisance for
   cluster members that are essentially passive receivers within a given
   multicast group. It is just as likely indicate that a passive member, with
   categroy has no
   outgoing VC already established to the group, will decide to send an
   IGMP reply value specified. Values in minutes are followed by
   'min', values in seconds are followed by 'sec'.

      Idle time for MARS - causing a VC MARS client pt to pt VC:
                                        [1 min, 20 min, -]

      Idle time for multipoint VCs from client.
                                        [1 min, 20 min, -]

      Allowed time between MARS_MULTI components.
                                        [-, -, 10 sec]

      Initial random L_MULTI_RQ/ADD retransmit timer range.
                                        [5 sec, -, 10 sec]

      Random time to be established were there was no need
   for one. This set VC_revalidate flag.
                                        [1 sec, -, 10 sec]

      MARS_JOIN/LEAVE retransmit interval.
                                        [5 sec, 10 sec, -]

      MARS_JOIN/LEAVE retransmit limit.
                                        [-, -, 5]

      Random time to re-register with MARS.
                                        [1 sec, -, 10 sec]

      Force wait if MARS re-registration is not a fatal problem looping.
                                        [1 min, -, -]

      Transmission interval for small clusters, but will
   seriously impact on the ability of a cluster MARS_REDIRECT_MAP.
                                        [1 min, 1 min, 2 min]

      Limit for client to scale.

   The most obvious solution is miss MARS_REDIRECT_MAPs.
                                        [-, -, 4 min]










Armitage                Expires April 24th, 1996                [Page 71]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


Appendix F.  Pseudo code for routers MARS operation.

   Implementations are entirely free to use comply with the
   MARS_GROUPLIST_REQUEST and MARS_GROUPLIST_REPLY messages, as
   described body of this
   memo in section 8.5. any way they see fit. This would remove appendix is purely for possible
   clarification.

   A MARS implementation might be built along the regular IGMP Queries,
   resulting lines suggested in cluster members only sending an IGMP Report when they
   first join
   this pseudo-code.

   1. Main

    1.1 Initilization

         Define a group.

   Alternative solutions do exist. One would be to modify the IGMP reply
   algorithm, for example:

      If the group member has VC open to the group proceed server list as per RFC
      1112 (picking a random reply delay between 0 and 10 seconds).

      If the group member does not have VC already open to the group,
      pick random reply delay between 10 and 20 seconds instead, and
      then proceed list of leaf nodes
                                            on ServerControlVC.
         Define a cluster list as per RFC 1112.

   If even one group member is sending to the group at the time list of leaf nodes
                                            on ClusterControlVC.
         Define a host map as the IGMP
   Query is issued then all list of hosts that are
                                            members of a group.
         Define a server map as the list of hosts (MCSs)
                                            that are serving a group.
         Read config file.
         Allocate message queues.
         Allocate internal tables.
         Set up passive receivers will find the IGMP
   Reply has been transmitted before their delay expires, so no new open VC
   is required. connection.
         Set up redirect_map timer.
         Establish logging.

    1.2 Message Processing

         Forever {
           If all group members are passive at the time of the IGMP
   Query then message has a response will eventually arrive, but 10 seconds later
   than under conventional circumstances.




Armitage                Expires March 20th, 1996                [Page 62]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th, 1995


   The preceeding solution requires re-writing existing IGMP code, and
   implies the ability of TLV then {
             If TLV is unsupported then {
               process as defined in TLV type field.
             } /* unknown TLV */
           } /* TLV present */
           Place incoming message in the IGMP entity to ascertain queue.
           For (all messages in the status of VCs
   on queue) {
             If the underlying ATM interface. This message is not likely to be available
   in a JOIN/LEAVE/MSERV/UNSERV with
               ar$flags.register == 1 then {
               If the short term.

   One short term solution message source is to provide something like the preceeding
   functionality with (not a 'hack' at the IP/ATM driver level within member of server list) &&
                (not a member of cluster
   members. Arrange for list) then {
                Drop the IP/ATM driver to snoop inside IP packets
   looking for IGMP traffic. message silently.
              }
             }
             If an IGMP packet is accepted for
   transmission, the IP/ATM driver can buffer it locally if there (ar$pro.type is no
   VC already active to that group. A 10 second timer not supported) or
                (the ATM source address is started, and if missing) then {
                Continue.



Armitage                Expires April 24th, 1996                [Page 72]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


             }
             Determine type of message.
             If an IGMP Reply for that group is received from elsewhere ERR_L_RELEASE arrives on ClusterControlVC then {
               Remove the
   cluster endpoints ATM address from all groups
               for which it has joined.
               Release the timer is reset. CMI.
               Continue.
             } /* error on CCVC */
             Call specific message handling routine.
             If the redirect_map timer expires, pops {
               Call MARS_REDIRECT_MAP message handling routine.
             } /* redirect timer pop */
           } /* all msgs in the IP/ATM driver
   then establishes a VC to queue */
         } /* forever loop */

   2. Message Handler

    2.1 Messages:

       - MARS_REQUEST

         Indicate no MARS_MULTI support of TLV.
         If the group supported TLV is not NULL then {
           Indicate MARS_MULTI support of TLV.
           Process as it would for a normal IP
   multicast packet.

   Some network implementors may find it advantageous required.
         } else { /* TLV NULL */
            Indicate message to configure be sent on Private VC.
            If the message source is a
   multicast member of server to support list then {
              If the group 224.0.0.1, rather than rely on
   a mesh. Given that IP multicast routers regularly send IGMP queries
   to this address, has a mesh will mean that each router will permanently
   consume an AAL context within each cluster member. In clusters served
   by multiple routers non-null host map then {
                Call MARS_MULTI with the VC load within switches in host map for the underlying ATM
   network will become a scaling problem.

   Finally, if a multicast server group.
              } else { /* no group */
                 Call MARS_NAK message routine.
              } /* no group */
            } else { /* source is used to support 224.0.0.1, another
   ATM driver level hack becomes a possible solution to IGMP Reply
   traffic.  The ATM driver may choose to grab all outgoing IGMP packets
   and send them out on cluster list */
               If the group has a non-null server map then {
                 Call MARS_MULTI with the VC established server map for sending to 224.0.0.1,
   regardless of the Class D address group.
               } else { /* cluster member but no server map */
                  If the IGMP message was actually for.
   Given that all hosts and routers must be members of 224.0.0.1, group has a non-null host map then {
                    Call MARS_MULTI with the
   intended recipients will still receive host map for the IGMP Replies. The negative
   impact group.
                  } else { /* no group */
                     Call MARS_NAK message routine.
                  } /* no group */
                 } /* cluster member but no server map */
              } /* source is that all a cluster members will receive the IGMP Replies. list */
            } /* TLV NULL */
         If a message exists then {
           Send message as indicated.
         }



Armitage                Expires March 20th, April 24th, 1996                [Page 63] 73]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


Appendix C.   Further comments on 'Clusters'.

   The cluster concept was introduced in section 1


         Return.

       - MARS_MULTI

         Construct a MARS_MULTI for two reasons.  The
   more well known term of Logical IP Subnet is both very IP specific,
   and constrained to unicast routing boundaries. As the architecture
   described in this document may be re-used in non-IP environments a
   more neutral term was needed. As specified map.
         If the needs of multicasting are not
   always bound by param indicates TLV support then {
           Process the same scopes TLV as unicasting, it was not immediately
   obvious that apriori limiting ourselves to LISs was beneficial in required.
         }
         Return.

       - MARS_JOIN

         If (ar$flags.copy != 0) silently ignore the
   long term.

   It must message.
         If more than a single <min,max> pair is specified then
         ignore the 2nd and subsequent pairs.
         Indicate message to be stressed that Clusters are purely an administrative being.
   You choose their size (i.e. sent on private VC.
         If (ar$flags.register == 1) then {
           If the number node is already a registered member of endpoints that register
   with the same MARS) based on your multicasting needs, and cluster
           associated with protocol type then { /*previous register*/
             Copy the
   resource consumption you are willing to put up with. The larger existing CMI into the
   number of ATM attached hosts you require multicast support for, MARS_JOIN.
           } else { /* new register */
              Add the
   more individual clusters you might choose node to establish (along with
   multicast routers ClusterControlVC.
              Add the node to provide inter-cluster traffic paths).

   Given that cluster list.
              ar$cmi = obtain CMI.
             } /* new register */
         } else { /* not all a register */
            If the hosts message source is in server map then {
              Drop the message silently.
              Indicate no message to be sent.
            } else {
               If the first <min,max> encompasses any given LIS may require multicast
   support, it becomes conceivable that you might assign group with
                                                 a single MARS server map then {
                 Call the Modified JOIN/LEAVE Processing routine.
                 Indicate no message to support hosts from across multiple LISs. In effect you have be sent.
               } else { /* server map does not exist */
                  Update internal tables.
                  Indicate message to be sent on ClusterControlVC.
                 } /* server map does not exist */
               }
              } /* not a
   cluster covering multiple LISs, and have achieved 'cut through'
   routing register */
         ar$flags.copy = 1.
         Send message as indicated.
         Return.

      - MARS_LEAVE

         If (ar$flags.copy != 0) silently ignore the message.
         Indicate message to be sent on ClusterControlVC.



Armitage                Expires April 24th, 1996                [Page 74]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


         If group address is 224.0.0.1 then {
           All references to this node must be eliminated from
            any other groups for multicast traffic. Under these circumstances increasing which it is a member.
         } /* group is 224.0.0.1 */
         If (ar$flags.register == 1) then { /* deregistration */
           Update internal tables to remove the geographical size of a member's ATM addr
           from all groups it has joined.
           Drop the endpoint from ClusterControlVC.
           Drop the endpoint from cluster might list.
           Release the CMI.
           Indicate message to be considered sent on Private VC.
         } else { /* not a good thing.

   However, practical considerations limit deregistration */
            If the size of clusters.  Having first <min,max> encompasses any group with
                                              a cluster span multiple LISs may server map then {
              Call the Modified JOIN/LEAVE Processing routine.
              Indicate no message to be sent.
            } else { /* server map does not always exist */
               Update internal tables.
               Indicate message to be sent on ClusterControlVC.
            }
         } /* not a particular 'win'
   situation.  As deregistration */
         If a message exists then {
           ar$flags.copy = 1.
           Send message as indicated.
         }
         Return.

    - MARS_MSERV

      If (ar$flags.register == 1) then { /* server register */
        Add the number of multicast capable hosts in your LISs
   increases it becomes more likely that you'll want endpoint as a leaf node to constrain ServerControlVC.
        Add the endpoint to the server list.
        Indicate the message to be sent on Private VC.
        ar$cmi = 0.
      } else { /* not a
   cluster's size register */
         If (group has non-null host map) && (no server map) then {
           Silently drop the message.
           Indicate no message to be sent.
         } else {
            If the source has not registered then {
              Drop and force multicast traffic ignore the message.
              Indicate no message to aggregate at multicast
   routers scattered across your be sent.
            } else {  /* source is registered */
               Add the server ATM cloud.

   Finally, multi-LIS clusters require addr to the server map for the group.
               Indicate the message to be sent on ServerControlVC.
               Send message as indicated.
               Make a degree copy of care when deploying
   IP multicast routers. Under the Classical IP model you need unicast
   routers message.
               Change the op code to MARS_JOIN.



Armitage                Expires April 24th, 1996                [Page 75]

Internet Draft       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


               Indicate the message to be sent on ClusterControlVC.
               ar$flags.layer3grp = 0.
               ar$flags.copy = 1.
              } /* source is registered */
           }
        } /* not a register */
      If a message exists then {
        Send message as indicated.
      }
      Return.

    - MARS_UNSERV

      If (ar$flags.register == 1) then { /* deregister */
        Remove the edges ATM addr of LISs. Under the MARS architecture you only
   need multicast routers at MCS from all server maps.
        If a server map becomes null then delete it.
        Remove the edges endpoint as a leaf of clusters. ServerControlVC.
        Remove the endpoint from server list.
        Indicate the message to be sent on Private VC.
      } else { /* not a deregister */
         If your cluster
   spans multiple LISs, the source is not a member of server list then {
          Drop and ignore the multicast routers will perceive
   themselves message.
          Indicate no message to have be sent.
         } else {  /* source is registered */
            Remove ATM addr of the MCS from each server map indicated.
            If a single interface that server map is simultaneously attached null then delete it.
            Indicate the message to multiple unicast subnets. Whether this situation will work depends be sent on ServerControlVC.
            Send message as indicated.
            Make a copy of the inter-domain multicast routing protocols you use, and your
   multicast router's ability to understand message.
            Change the new relationship between
   unicast and multicast topologies.

   In op code to MARS_LEAVE.
            Indicate the absence of futher research in this area, networks deployed in
   conformance message (copy) to this document MUST make their IP cluster and IP LIS
   coincide, so be sent on ClusterControlVC.
            ar$flags.layer3grp = 0;
            ar$flags.copy = 1.
           } /* source is registered */
        } /* not a deregister */
      If a message exists then {
        Send message as to avoid these complications. indicated.
      }
      Return.

    - MARS_NAK

      Build command.
      Return.

    - MARS_GROUPLIST_REQUEST

      If (ar$pnum != 1) then Return.



Armitage                Expires March 20th, April 24th, 1996                [Page 64] 76]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


Appendix D.  TLV list parsing algorithm.

   The following pseudo-code represents how


      Call MARS_GROUPLIST_REPLY with the TLV list format
   described range and output VC.
      Return.

    - MARS_GROUPLIST_REPLY

      Build command for specified range.
      Indicate message to be sent on specified VC.
      Send message as indicated.
      Return.

    - MARS_REDIRECT_MAP

       Include the MARSs own address in section 10 could the message.
       If there are backup MARSs then include their addresses.
       Indicate MARS_REDIRECT_MAP is to be handled by a MARS or MARS client.

      list = (ar$extoff & 0xFFFC);

      if (list == 0) exit;

      list = list + message_base;

      while (list->Type.y != 0)
            {
                  switch (list->Type.y)
                  {
                        default: sent on ClusterControlVC.
       Send message back as indicated.
       Return.

   3. Send Message Handler

      If the message is going out ClusterControlVC then {
                           if (list->Type.x == 0) break;

                           if (list->Type.x == 1) exit;

                           if (list->Type.x == 2) log-error-and-exit;
                          }

                        [...other handling goes here..]
       ar$msn = obtain a CSN
      }

                  list += (list->Length + 4 + ((4-(list->Length & 3)) %
                  4));
      If the message is going out ServerControlVC then {
       ar$msn = obtain a SSN
      }

      return;
      Return.

   4.  Number Generator

   4.1 Cluster Sequence Number

       Generate the next sequence number.
       Return.

   4.2 Server Sequence Number

       Generate the next sequence number.
       Return.

   4.3 CMI

       CMIs are allocated uniquely per registered cluster member
       within the context of a particular layer 3 protocol type.
       A single node may register multiple times if it supports
       multiple layer 3 protocols.
       The CMIs allocated for each such registration may or may
       not be the same.



Armitage                Expires March 20th, April 24th, 1996                [Page 65] 77]

Internet Draft       <draft-ietf-ipatm-ipmc-07.txt> September 20th,       <draft-ietf-ipatm-ipmc-08.txt>   October 24th, 1995


Appendix E.  Summary of timer values.


       Generate a CMI for this protocol.
       Return.

   5. Modified JOIN/LEAVE Processing

      This appendix summarises of various timers or limits mentioned in the
   main body routine processes JOIN/LEAVE when a server map exists.

      Make a copy of the document. Values are specified in message.
      Change the following
   format:  [x, y, z] indicating a minimum value of x, a recommended
   value type of y, and the copy to MARS_SJOIN.
      If the message is a maximum value MARS_LEAVE then {
       Change the type of z. A '-' will indicate that a
   categroy has no value specified. Values the copy to MARS_SLEAVE.
      }
      ar$flags.copy = 1 (copy).
      Indicate the message to be sent on ServerControlVC.
      Send message (copy) as indicated.
      ar$flags.punched = 0 in minutes are followed the original message.
      Indicate the message to be sent on Private VC.
      Send message (original) as indicated.
      Hole punch the <min,max> group by
   'min', values in seconds excluding
        from the range those groups that are followed served by 'sec'.

      Idle time for MARS - MARS client pt to pt VC:
                                        [1 min, 20 min, -]

      Idle time for multipoint VCs from client.
                                        [1 min, 20 min, -]

      Allowed time between MARS_MULTI components.
                                        [-, -, 10 sec]

      Initial random L_MULTI_RQ/ADD retransmit timer range.
                                        [5 sec, -, 10 sec]

      Random time to set VC_revalidate flag.
                                        [1 sec, -, 10 sec]

      MARS_JOIN/LEAVE retransmit interval.
                                        [5 sec, 10 sec, -]

      MARS_JOIN/LEAVE retransmit limit.
                                        [-, -, 5]

      Random time to re-register with MARS.
                                        [1 sec, -, 10 sec]

      Force wait if MARS re-registration is looping.
                                        [1 min, -, -]

      Transmission interval for MARS_REDIRECT_MAP.
                                        [1 min, 1 min, 2 min]

      Limit for client MCSs.
      Indicate the (original) message to miss MARS_REDIRECT_MAPs.
                                        [-, -, 4 min] be sent on ClusterControlVC.
      If (number of holes punched > 0) then { /* punched holes */
        In original message do {
         ar$flags.punched = 1.
         old punched list <- new punched list.
        }
      } /* punched holes */
      ar$flags.copy = 1.
      Send message as indicated.
      Return.





















Armitage                Expires March 20th, April 24th, 1996                [Page 66] 78]


----