wdiff draft-ietf-idmr-cbt-spec-06.txt draft-ietf-idmr-cbt-spec-07.txt

Inter-Domain Multicast Routing (IDMR) A. Ballardie
INTERNET-DRAFT University College London
S. Reeve & N. Jain
Bay Networks, Inc.

September 1996 Consultant

March 1997

Core Based Trees (CBT) Multicast Routing

-- Protocol Specification --

Status of this Memo

This document is an Internet Draft. Internet Drafts are working doc-
uments of the Internet Engineering Task Force (IETF), its Areas, and
its Working Groups. Note that other groups may also distribute work-
ing 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 I-D abstract listing contained in each Internet
Draft directory to learn the current status of this or any other
Internet Draft.

Abstract

This document describes the Core Based Tree (CBT) network layer mul-
ticast routing protocol. CBT is builds a next-generation shared multicast distribution
tree per group, and is suited to inter- and intra-domain multicast
routing.

CBT is protocol independent in that it makes use of a shared delivery tree rather than separate per-sender
trees utilized by most other multicast schemes [1, 2, 3]. unicast routing
to establish paths between senders and receivers. The CBT
architecture architec-
ture is described in [4a].

This specification includes an optimization whereby unencapsulated
(native) IP-style multicasts are forwarded by CBT routers, resulting
in very good forwarding performance. This mode of operation is
called CBT "native mode". Native mode can only be used in CBT-only
domains (footnote 1).
_________________________
This revision contains two appendices; Appendix A describes simple
CBT add-on mechanisms for dynamically migrating a CBT tree to one
whose core is directly attached to a source's subnetwork, thereby
allowing CBT to emulate shortest-path trees. Appendix B describes a
group state aggregation scheme. [1].

This document is progressing through the IDMR working group of the
IETF. CBT related documents include [4, 5]. [1, 5, 6]. For all IDMR-related
documents, see http://www.cs.ucl.ac.uk/ietf/idmr.

NOTE that core placement

TABLE OF CONTENTS

1. Changes Since Previous Revision............................ 3

2. Introduction & Terminology................................. 4

3. CBT Functional Overview.................................... 5

4. CBT Protocol Specificiation Details........................ 8

4.1 CBT HELLO Protocol..................................... 8

4.1.1 Sending HELLOs................................... 9

4.1.2 Receiving HELLOs................................. 9

4.2 JOIN_REQUEST Processing................................ 10

4.2.1 Sending JOIN_REQUESTs............................ 10

4.2.2 Receiving JOIN_REQUESTs.......................... 10

4.3 JOIN_ACK Processing.................................... 11

4.3.1 Sending JOIN_ACKs................................ 11

4.3.2 Receiving JOIN_ACKs.............................. 12

4.4 QUIT_NOTIFICATION Processing........................... 12

4.4.1 Sending QUIT_NOTIFICATIONs....................... 12

4.4.2 Receiving QUIT_NOTIFICATIONs..................... 13

4.5 CBT ECHO_REQUEST Processing............................ 14

4.5.1 Sending ECHO_REQUESTs............................ 14

4.5.2 Receiving ECHO_REQUESTs.......................... 14

4.6 ECHO_REPLY Processing.................................. 15

4.6.1 Sending ECHO_REPLYs.............................. 15

4.6.2 Receiving ECHO_REPLYs............................ 15
4.7 FLUSH_TREE Processing.................................. 16

4.7.1 Sending FLUSH_TREE Messages...................... 16

4.7.2 Receiving FLUSH_TREE Messages.................... 16

5. Timers and management is not discussed in this doc-
ument. Default Values.................................. 16

6. CBT Packet Formats and Message Types....................... 17

6.1 CBT Common Control Packet Header....................... 18

6.2 HELLO Packet Format.................................... 19

6.3 JOIN_REQUEST Packet Format............................. 19

6.4 JOIN_ACK Packet Format................................. 20

6.5 QUIT_NOTIFICATION Packet Format........................ 21

6.6 ECHO_REQUEST Packet Format............................. 21

6.7 ECHO_REPLY Packet Format............................... 22

6.8 FLUSH_TREE Packet Format............................... 23

7. Core Router Discovery...................................... 23

7.1 Bootstrap Message Format.............................. 25

7.2 Candidate Core Advertisement Message Format........... 25

8. Interoperability Issues.................................... 25

Acknowledgements.............................................. 26

References.................................................... 26

Author Information............................................ 27

1. Changes since Previous Revision (05)

This note summarizes the changes to this document since the previous revision (revision 05).

+o inclusion of "first hop router" and "primary core" fields in the CBT mode data packet header.

+o removal of the term "non-core" router, replaced by "on-tree"
router.

+o removal of protocol specification differs significantly
from the term "default DR (D-DR)", replaced simply by DR.

+o inclusion previously released revision (05). Consequently, this revi-
sion represents version 2 of T and S bits in the CBT control and data packet
headers (type of service, protocol. CBT version 2 is not,
and security, respectively).

+o was not, intended to be backwards compatible with version 1; we
do not expect this to cause extensive compatibility problems because
we do not believe CBT control messages are now carried directly over IP rather
than UDP (for is at all implementations).

+o inclusion widely deployed at this stage. How-
ever, any future versions of an Appendix (A) describing extensions to the CBT
protocol can be expected to achieve dynamic source-migration of core routers for
shortest-path tree emulation. be backwards com-
patible with this version.

The most significant changes to version 2 compared to version 1
include:

+o inclusion new LAN mechanisms, including the incorporation of an Appendix (B) describing a group state aggrega-
tion scheme.
_________________________
1 The term "domain" should be considered synonymous HELLO pro-
tocol.

+o new simplified packet formats, with "routing domain" throughout, as are the terms "re-
gion" and "cloud". definition of a common
CBT control packet header.

+o editorial changes a generic intra-domain core discovery ("bootstrap") mechanism,
to be specified separately, and some re-organisation throughout for extra
clarity. published soon.

This specification revision is a complete re-write of the previous
revision.

2. Some Introduction & Terminology

In CBT, the core routers for a particular group are categorised into
PRIMARY CORE, and NON-PRIMARY (secondary) CORES.

The "core tree" router" (or just "core") is the part of a tree linking all core routers of router which configured
to act as a
particular "meeting point" between a sender and group together.

On-tree routers are those with receivers. The
term "rendezvous point (RP)" is used equivalently in some contexts
[2]. Each core router is configured to know it is a forwarding database entry for the
corresponding group.

3. Protocol Specification

3.1. Tree Joining Process -- Overview core router.

A CBT router that is notified part of a local host's desire to join a group via
IGMP [6]. We refer to a CBT router with directly attached hosts distribution tree is known as a
"leaf CBT router", or just "leaf" router.

The following CBT control messages come into play subequent to a sub-
net's CBT leaf router receiving an IGMP membership report (also
termed "IGMP join"):

+o JOIN_REQUEST

+o JOIN_ACK

If the CBT leaf "on-
tree" router. An on-tree router is maintains active state for the subnet's designated router (see next
section), it generates a CBT join-request in response group.

We refer to receiving an
IGMP group membership report from a directly connected host. The CBT
join broadcast interface as any interface that supports mul-
ticast transmission.

An "upstream" interface (or router) is one which is sent to the next-hop on the unicast path to a target core,
specified in
towards the join packet; a group's core router elects a "target core" based
on a static configuration. If, with respect to this router. A "down-
stream" interface (or router) is one which is on receipt of an IGMP-join, the
locally-elected DR has already joined path away from
the corresponding tree, then it
need do nothing more group's core router with respect to joining. this router.

Other terminology is introduced in its context throughout the text.

3. CBT Functional Overview

The join CBT protocol is processed designed to build and maintain a shared multicast
distribution tree that spans only those networks and links leading to
interested receivers.

To achieve this, a host first expresses its interest in joining a
group by each such multicasting an IGMP host membership report [3] across its
attached link. On receiving this report, a local CBT aware router
invokes the tree joining process (unless it has already) by generat-
ing a JOIN_REQUEST message, which is sent to the next hop on the path to
towards the core, until
either group's core router (how the local router discovers which
core to join reaches is discussed in section 7). This join message must be
explicitly acknowledged (JOIN_ACK) either by the target core router itself,
or hits a by another router that is on the unicast path between the sending
router and the core, which itself has already part successfully joined the
tree.

The join message sets up transient join state in the routers it tra-
verses, and this state consists of <group, incoming interface, outgo-
ing interface>. "Incoming interface" and "outgoing interface" may be
"previous hop" and "next hop", respectively, if the corresponding distribution tree (as identified
by
links do not support multicast transmission. "Previous hop" is taken
from the group address). In both cases, incoming control packet's IP source address, and "next hop"
is gleaned from the router concerned terminates routing table - the join, and responds next hop to the specified
core address. This transient state eventually times out unless it is
"confirmed" with a join-ack (join acknowledgement), which join acknowledgement (JOIN_ACK) from upstream. The
JOIN_ACK traverses the reverse-path reverse path of the corresponding join. This join mes-
sage, which is possi-
ble possible due to the presence of the transient path state created by a join traversing a
CBT router. The ack fixes that
state.

3.2. DR Election

Multiple CBT routers may be connected Once the acknowledgement reaches the router that originated
the join message, the new receiver can receive traffic sent to the
group.

Loops cannot be created in a multi-access subnetwork.
In such cases it CBT tree because a) there is necessary only one
active core per group, and b) tree building/maintenance scenarios
which may lead to elect the creation of tree loops are avoided. For exam-
ple, if a subnetwork designated router's upstream neighbour becomes unreachable, the router
(DR) that is responsible for generating and sending CBT joins
upstream, on behalf
immediately "flushes" all of hosts its downstream branches, allowing them
to individually rejoin if necessary. Transient unicast loops do not
pose a threat because a new join message that loops back on itself
will never get acknowledged, and thus eventually times out.

The state created in routers by the subnetwork.

CBT DR election happens "on the back" sending or receiving of IGMP [6]; on a subnet with
multiple multicast routers, an IGMP "querier"
JOIN_ACK is elected as part bi-directional - data can flow either way along a tree
"branch", and the state is group specific - it consists of
IGMP. At start-up, the group
address and a multicast router assumes list of local interfaces over which join messages for
the group have previously been acknowledged. There is no other multicast
routers are present on its subnetwork, and so begins by believing concept of
"incoming" or "outgoing" interfaces, though it is necessary to be
able to distinguish the subnet's IGMP querier. It sends upstream interface from any downstream inter-
faces. In CBT, these interfaces are known as the "parent" and "child"
interfaces, respectively. We recommend the parent be distinguished as
such by a small number IGMP-HOST-
MEMBERSHIP-QUERYs in short succession single bit in order each multicast forwarding cache entry.

With regards to quickly learn about
any group memberships on the subnet. If other information contained in the multicast routers are
present forwarding
cache, on the same subnet, they will receive these IGMP queries; a link types not supporting native multicast transmission an
on-tree router yields querier duty as soon as it hears an IGMP
query from a lower-addressed router on the same subnetwork.

The CBT DR is always must store the subnet's IGMP querier (footnote 2). As address of a
result, there parent and any children.
On links supporting multicast however, parent and any child informa-
tion is no protocol overhead whatsoever associated represented with
electing a CBT D-DR.

3.3. Tree Joining Process -- Details

The receipt of local interface addresses (or similar iden-
tifying information, such as an IGMP group membership report by interface "index") over which the
parent or child is reachable.

When a CBT DR for multicast data packet arrives at a CBT
group not previously heard from triggers router, the tree joining process; router uses the DR unicasts a JOIN-REQUEST to
group address as an index into the first hop on multicast forwarding cache. A copy
of the (unicast) path incoming multicast data packet is forwarded over each inter-
face (or to the target core specified each address) listed in the CBT join packet.

_________________________
2 Or lowest addressed CBT router if entry except the subnet's IGMP
querier is non-CBT capable. incoming
interface.

Each CBT-capable router traversed on the path between the sending DR
and that comprises a CBT multicast tree, except the core processes
router, is responsible for maintaining its upstream link, provided it
has interested downstream receivers, i.e. the join. However, if child interface list is
non-NULL. A child interface is one over which a join hits member host is
directly attached, or one over which a CBT downstream on-tree router
that is already on-tree, the join
attached. This "tree maintenance" is not propogated further, but
acknowledged achieved by each downstream from that point.

JOIN-REQUESTs carry the identity of all
router periodically sending a CBT "keepalive" message (ECHO_REQUEST)
to its upstream neighbour, i.e. its parent router on the cores associated tree. One
keepalive message is sent to represent entries with the
group. Assuming there same parent,
thereby improving scalability on links which are no on-tree routers in between, once the
join (subcode ACTIVE_JOIN) reaches shared by many
groups. On multicast capable links, a keepalive is multicast to the target core, if
"all-cbt-routers" group (IANA assigned as 224.0.0.15); this has a
suppressing effect on any other router for which the target
core link is not the primary core (as indicated in its par-
ent link. If a separate field of the
join packet) it first acknowledges the received join by means parent link does not support multicast transmission,
keepalives are unicast.

The receipt of a
JOIN-ACK, then sends keepalive message over a JOIN-REQUEST, subcode REJOIN-ACTIVE, to the
primary core router.

If the rejoin-active reaches the primary core, it responds by sending valid child interface imme-
diately prompts a JOIN-ACK, subcode PRIMARY-REJOIN-ACK, response (ECHO_REPLY), which traverses the reverse-
path of the join (rejoin). The primary-rejoin-ack serves to confirm
no loop is present, either unicast or
multicast, as appropriate.

The ECHO_REQUEST does not contain any group information; the
ECHO_REPLY does, but only periodically. To maintain consistent infor-
mation between parent and so explicit loop detection child,
the parent periodically reports, in an ECHO_REPLY, all groups for
which it has state, over each of its child interfaces for those
groups. This group-carrying echo reply is not necessary.

If some other on-tree router prompted explicitly by
the receipt of an echo request message. A child is encountered before notified of the rejoin-active
reaches
time to expect the primary, that router responds with next echo reply message containing group informa-
tion in an echo reply prompted by a JOIN-ACK, subcode
NORMAL. On receipt child's echo request. The fre-
quency of parent group reporting is at the ack, subcode normal, granularity of minutes.

It cannot be assumed all of the router sends routers on a
join, subcode REJOIN-NACTIVE, which acts as multi-access link have a loop detection packet
(see section 8.3). Note that loop detection
uniform view of unicast routing; this is not necessary subse-
quent particularly the case when a
multi-access link spans two or more unicast routing domains. This
could lead to receiving multiple upstream tree branches being formed (an error
condition) unless steps are taken to ensure all routers on the link
agree which is the upstream router for a join-ack with subcode PRIMARY-REJOIN-ACK.

To facilitate detailed protocol description, we use particular group. CBT
routers attached to a sample topol-
ogy, illustrated multi-access link participate in Figure 1 (shown over). Member hosts are shown an explicit
election mechanism that elects a single router, the designated router
(DR), as
individual capital letters, routers are prefixed with R, and subnets
are prefixed with S.

A B
| S1 S4 |
------------------- -----------------------------------------------
| | | |
------ ------ ------ ------
| R1 | | R2 | | R5 | | R6 |
------ ------ ------ ------
C | | | | |
| | | | S2 | S8 |
---------- ------------------------------------------ -------------
S3 |
------
| R3 |
| ------ D
| S9 | | S5 |
| | ---------------------------------------------
| |----| | |
---| R7 |-----| ------
| |----| |------------------| R4 |
| S7 | ------ F
| | | S6 |
|-E | ---------------------------------
| |
| ------
|---| |---------------------| R8 |
|R12 ----| ------ G
|---| | | | S10
| S14 ----------------------------
| |
I --| ------
| | R9 |
------
| S12
| ----------------------------
S15 | |
| ------
|----------------------|R10 |
J ---| ------ H
| | |
| ----------------------------
| S13

Figure 1. Example Network Topology
Taking the example topology link's upstream router for all groups. Since the DR
might not be the link's best next-hop for a particular core router,
this may result in figure 1, host A wishes to join group
G. All subnets' routers have been configured to use core routers R4
(primary core) and R9 (secondary core) for messages being re-directed back across a range of group
addresses, including G.

Router R1 receives an IGMP host membership report, and proceeds to
multi-access link. If this happens, the re-directed join message is
unicast a JOIN-REQUEST, subcode ACTIVE-JOIN to across the next-hop on link by the
path DR to R4 (R3), the target core. R3 receives the join, caches the
necessary group information (transient state), and forwards it best next-hop, thereby pre-
venting a looping scenario. This re-direction only ever applies to R4
-- the target of the join.

R4, being the target of the join, sends
join messages. Whilst this is suboptimal for join messages, which
are generated infrequently, multicast data never traverses a JOIN_ACK (subcode NORMAL)
back out of link
more than once (either natively, or encapsulated).

In all but the receiving interface exception case described above, all CBT control mes-
sages are multicast over multicast supporting links to the previous-hop sender "all-cbt-
routers" group, with IP TTL 1. The IP source address of the
join, R3. A JOIN-ACK, like a JOIN-REQUEST, CBT control
messages is processed hop-by-hop by
each router on the reverse-path outgoing interface of the corresponding join. sending router. The
receipt IP des-
tination address of a join-ack establishes CBT control messages is either the receiving "all-cbt-
routers" group address, or the IP address of a router reachable over
one of the sending router's interfaces, depending on whether the corre-
sponding CBT tree, i.e.
sender's outgoing link supports multicast transmission. All the router becomes nec-
essary addressing information is obtained as part of tree set up.

If CBT is implemented over a branch on the
delivery tree. Finally, R3 sends tunnelled topology, when sending a join-ack to R1. A new CBT branch
has been created, attaching subnet S1 to the CBT delivery tree for
the corresponding group.

For the period between any CBT-capable router forwarding (or origi-
nating) a JOIN_REQUEST and receiving
control packet over a JOIN_ACK tunnel interface, the corresponding sending router is not permitted to acknowledge any subsequent joins received
for uses as
the same group; rather, packet's IP source address the router caches such joins till such
time as it has itself received a JOIN_ACK for local tunnel end point address,
and the original join. Only
then can it acknowledge any cached joins. A router is said to be in a
"pending-join" state if it is awaiting a JOIN_ACK itself.

Note that remote tunnel end point address as the presence packet's IP destina-
tion address.

4. Protocol Specification Details

Details of asymmetric routes in the underlying unicast
routing does not affect the tree-building process; CBT tree branches
are symmetric by the nature in which they protocol are built. Joins set up
transient state (incoming and outgoing interface state) presented in all
routers along the context of a path single
router implementation.

4.1. CBT HELLO Protocol

The HELLO protocol is used to elect a particular core. The corresponding join-ack
traverses the reverse-path of the join designated router (DR) on
broadcast-type links. It is also used to elect a designated border
router (BR) when interconnecting a CBT domain with other domains (see
[5]).

A router represents its status as dictated a link's DR by setting the transient
state, and not necessarily the path DR-flag
on that underlying routing would
dictate. Whilst permanent asymmetric routes could pose interface; a problem for
CBT, transient asymmetricity DR flag is detected by associated with each of a router's
broadcast interfaces. This flag can only assume one of two values:
TRUE or FALSE. By default, this flag is FALSE.

HELLO messages are multicast periodically to the CBT protocol.

3.4. Forwarding Joins on Multi-Access Subnets all-cbt-routers
group, 224.0.0.15, using IP TTL 1. The DR election mechanism does not guarantee that advertisement period is
[HELLO_TIMER] seconds. [HELLO_TIMER] comprises a configured
[HELLO_INTERVAL], to which is added [RND_RSP] seconds - a random
response interval. This random response additive is required to
avoid the DR will be potential problem of synchronisation between HELLO adver-
tisements (or other control messages) from different routers. The
HELLO protocol's convergence time is set at [HELLO_CONV] seconds -
the time after which no further HELLOs are expected in any one round
of the protocol.

Each HELLO advertising router includes the upper bound of its
[RND_RSP] timer in its HELLO advertisements. This is necessary so
that actually forwards a join off a multi-access network; all routers attached to the
first hop link can agree on a common HELLO
convergence time [HELLO_CONV]; in any one round of the path to HELLO proto-
col, a particular core might be via another router on the same subnetwork, which actually forwards off-subnet.

Although very much assumes the same, let's see another example using our
example topology of figure 1 minimum of a host joining a CBT tree for the
case where more than one CBT router exists on the host subnetwork.

B's subnet, S4, has 3 CBT routers attached. Assume also that R6 has
been elected IGMP-querier and CBT DR.

R6 (S4's DR) receives an IGMP group membership report. R6's upper bound of its config-
ured information suggests R4 [RND_RSP] and that of any received advertisement's. The minimum
upper bound is then used as the target core for this group. R6
thus generates a join-request for target core R4, subcode
ACTIVE_JOIN. R6's routing table says the next-hop on router's [RND_RSP] upper bound in
the path to R4
is R2, which is on next round of the same subnet as R6. This protocol. [HELLO_CONV] is irrelevant to R6,
which unicasts it to R2. R2 unicasts it to R3, which happens set to be
already on-tree this minimum
upper bound + 2 seconds (the 2 seconds being a response "safety mar-
gin") for the specified group (from R1's join). R3 there-
fore can acknowledge the arrived join and unicast the ack back to R2.
R2 forwards it to R6, the origin next round of the join-request.

If an IGMP membership report is received by a DR with protocol.

A network manager can preference a join for the
same group already pending, or if the router's DR is already on-tree for eligibility by option-
ally configuring a HELLO preference. Valid configuration values range
from 1 to 254 (decimal), 1 representing the
group, it takes no action.

3.5. On-Demand "Core Tree" Building

The "core tree" - "most eligible" value. In
the part absence of explicit configuration, a CBT tree linking all router assumes the default
HELLO preference value of its cores
together, 255. The elected DR uses HELLO preference
zero (0) in HELLO advertisements, irrespective of any configured
preference. The DR continues to use preference zero for as long as
it is built on-demand. That is, the core tree running.

The DR election winner is only built
subsequent to a non-primary (secondary) core receiving a join-
request. This triggers that which advertises the secondary core to join lowest HELLO
preference, or the primary core; lowest-addressed in the primary need never join anything.

Join-requests carry an list event of core a tie.

The situation where two or more routers (and attached to the identity of same broad-
cast link are advertising HELLO preference 0 should never arise. How-
ever, should this situation arise, all but the
primary core in lowest addressed zero-
advertising router relinquishes its own separate field), making it possible for the
secondary cores to know where to join when they themselves receive a
join. Hence, the primary core must be uniquely identified claim as such
across the whole group. A secondary joins DR immediately by unset-
ting the primary subsequent to
sending an ack for DR flag on the first join corresponding interface. The relinquishing
router(s) subsequently advertise their previously used preference
value in HELLO advertisements.

4.1.1. Sending HELLOs

When a router starts up, it receives.

3.6. Tree Teardown

There are multicasts two scenarios whereby HELLO messages over each
of its broadcast interfaces in successsion. The DR flag is initially
unset (FALSE) on each broadcast interface.

A router sends a tree branch may be torn down:

+o During HELLO message whenever its [HELLO_TIMER] expires.

Whenever a re-configuration. If router sends a router's best next-hop to the
specified core is one HELLO message, it resets its [HELLO_TIMER].

4.1.2. Receiving HELLOs

On receipt of any HELLO message, a router adjusts its existing children, then before
sending [RND_RSP] upper
bound to the join it must tear down minimum of this router's configured [RND_RSP] upper
bound and that particular downstream
branch. It does so by sending a FLUSH_TREE message which is pro-
cessed hop-by-hop down received in the branch. All routers receiving this
message must process it and forward it to all their children.
Routers that have received HELLO. The router also
adjusts its [HELLO_CONV] as described above.

A router need not respond to a flush HELLO message will re-establish
themselves on the delivery tree if they have directly connected
subnets with group presence.

+o If a CBT router has no children it periodically checks all its
directly connected subnets for group member presence. If no mem-
ber presence the received HELLO is ascertained on any of
"better" than its subnets it sends a
QUIT_REQUEST upstream to remove itself from own. Thus, in steady state, the tree.

The receipt of a quit-request triggers HELLO protocol
incurs very little traffic overhead.

If the receiving parent
router to received HELLO message is "better" (lower preferenced, or
equally preferenced but lower addressed) than it would send itself,
it immediately query unsets its forwarding database to establish
whether there remains any directly connected group membership,
or any children, for the said group. If not, DR flag on the router itself
sends a quit-request upstream.

The following example, using arriving interface if the example topology of figure 1, shows
how a tree branch DR
flag is gracefully torn down using a QUIT_REQUEST.

Assume group member B leaves group G set on subnet S4. B issues an IGMP
HOST-MEMBERSHIP-LEAVE (relevant only to IGMPv2 and later versions) that interface. It also resets its [HELLO_TIMER].

If the received HELLO message which is multicast to the "all-routers" group (224.0.0.2).
R6, not "better" than this router would
send itself, it sets its [RND_RSP] random response timer; on expiry,
the subnet's DR and IGMP-querier, router responds with a group-specific-
QUERY. No hosts respond its own HELLO message . If no "better" HELLO
message is received within the required response interval, so current [HELLO_CONV], the router sets
the DR
assumes group G traffic is no longer wanted flag on subnet S4.

Since R6 has no the corresponding interface.

4.2. JOIN_REQUEST Processing

A JOIN_REQUEST is the CBT children, and no other directly attached subnets
with group G presence, it immediately follows on by sending a
QUIT_REQUEST control message used to R2, its parent on register a member
host's interest in joining the distribution tree for group G. R2 responds
with a QUIT-ACK, unicast to R6; R2 removes the corresponding child
information. R2 in turn sends group.

4.2.1. Sending JOIN_REQUESTs

A JOIN_REQUEST can only ever be originated by a QUIT upstream to R3 (since it has no
other children or subnet(s) with group presence).

NOTE: immediately subsequent to sending leaf router, i.e. a QUIT-REQUEST, the sender
removes
router with directly attached member hosts. This join message is sent
hop-by-hop towards the corresponding parent information, i.e. it does not
wait core router for the receipt group (see section 7).
The originating router caches <group, NULL, upstream interface> state
for each join it originates. This state is known as "transient join
state". The absence of a QUIT-ACK.

R3 responds to "downstream interface" (NULL) indicates
that this router is the QUIT by unicasting a QUIT-ACK to R2. R3 subse-
quently checks whether it in turn can send a quit by checking group G
presence on its directly attached subnets, join message originator, and is therefore
responsible for any group G children. retransmissions of this message if a response is
not received within [JOIN_RTX_INTERVAL]. It has the latter (R1 is its child on an error if no
response is received after [JOIN_TIMEOUT] seconds. If this error
condition occurs, the group G tree), and so R3
cannot itself send a quit. However, joining process may be re-invoked by the branch R3-R2-R6 has been
removed from
receipt of the tree.

4. Tree Maintenance

Once a tree branch has been created, i.e. a CBT router has received next IGMP host membership report from a
JOIN_ACK for locally
attached member host.

Note that if the interface over which a JOIN_REQUEST previously sent (or forwarded), a child
router is required to monitor be sent
supports multicast, the status of its parent/parent link at
fixed intervals by means of a "keepalive" mechanism operating between
them. The "keepalive" protocol JOIN_REQUEST is simple, and implemented by means
of two CBT control messages: CBT_ECHO_REQUEST and CBT_ECHO_REPLY; a
child unicasts a CBT-ECHO-REQUEST multicast to its parent, which unicasts a
CBT-ECHO-REPLY in response.

Adjacent CBT the all-cbt-
routers only need to send one keepalive representing all
children having group, using IP TTL 1. If the same parent, reachable over a particular link,
regardless of group. This aggregation strategy link does not support multi-
cast, the JOIN_REQUEST is expected unicast to con-
serve considerable bandwidth the next hop on "busy" links, such as transit net-
work, or backbone network, links.

For any CBT router, if its parent router, or the unicast path
to the parent,
fails, group's core.

4.2.2. Receiving JOIN_REQUESTs

On broadcast links, JOIN_REQUESTs which are multicast may only be
forwarded by the child is initially responsible for re-attaching itself,
and therefore all link's DR. Other routers subordinate to it on the same branch, attached to the tree.

4.1. Router Failure

An on-tree router can detect a failure from the following two cases:

+o if the child responsible for sending keepalives across a partic-
ular link stops receiving CBT_ECHO_REPLY messages. In this case
the child realises that its parent has become unreachable and
must therefore try and re-connect to the tree for all groups
represented on may
process the parent/child link. For all groups sharing a
common core set (corelist), provided those groups can be speci-
fied as a CIDR-like aggregate, an aggregated join can be sent
representing the range of groups. Aggregated joins (see below). JOIN_REQUESTs which are made
possible multicast over
a point-to-point link are only processed by the presence of a "group mask" field in router on the CBT con-
trol packet header (footnote 3).

If link
which does not have a range of groups cannot local interface corresponding to the join's
network layer (IP) source address. Unicast JOIN_REQUESTs may only be represented
processed by a mask, then each
group must be re-joined individually.

CBT's re-join strategy is as follows: the rejoining router which
is immediately subordinate has a local interface corresponding to
the failure sends join's network layer (IP) destination address.

With regard to forwarding a JOIN_REQUEST
(subcode ACTIVE_JOIN received JOIN_REQUEST, if it has no children attached, the receiving
router is not on-tree for the group, and subcode
ACTIVE_REJOIN if at least one child is attached) not the group's core
router, the join is forwarded to the best
next-hop router next hop on the path to towards the elected
core. If no JOIN-ACK The join is received after three retransmissions, each transmission being
at PEND-JOIN-INTERVAL (5 secs) intervals, multicast, or unicast, according to whether the next-highest pri-
ority core is elected from
outgoing interface supports multicast. The router caches the core list, and follow-
ing information with respect to the process
repeated. If all cores have been tried unsuccessfully, the DR
has no option but to give up.

+o if a parent stops receiving CBT_ECHO_REQUESTs from a child. In forwarded join: <group, down-
stream interface, upstream interface>.

If this case, if the parent has transient join state is not received an expected keepalive
after CHILD_ASSERT_EXPIRE_TIME, all children reachable across
that link are removed "confirmed" with a join acknowl-
edgement (JOIN_ACK) message from upstream, the parent's forwarding database.

4.2. Router Re-Starts

There are two cases to consider here:

+o Core re-start. All JOIN-REQUESTs (all types) carry the identi-
ties (i.e. IP addresses) of each of the cores for a group. state is timed out
after 1.5 times [JOIN_RTX_INTERVAL].

If a the receiving router is a core for a group, but has only recently re-started,
it will not be aware that it is a core for any group(s). In such
circumstances, a the group's core only becomes aware that it router, the join is such "ter-
minated" and acknowledged by
receiving a JOIN-REQUEST. Subsequent to means of a core learning its
status in this way, JOIN_ACK. Similarly, if it is not the primary core it acknowl-
edges
router is on-tree and the received join, then sends a JOIN_REQUEST (subcode
ACTIVE_REJOIN) to arrives over an interface that
is not the primary core. If upstream interface for the re-started router is group, the primary core, it need take no action, i.e. in all
_________________________
3 There are situations where it join is advantageous acknowl-
edged.

If [RND_RSP] pertaining to
send a single join-request that represents potentially
many groups. One such example JOIN_REQUEST is provided in [11],
whereby active (i.e. running),
if a designated border router JOIN_REQUEST is required to join
all groups inside a CBT domain.

circumstances, received for the primary core simply waits to be joined by
other routers.

+o Non-core re-start. In this case, same group over that group's
parent interface, cancel [RND_RSP] for the impending JOIN_REQUEST.

If this router can only join has a cache-deletion-timer [CACHE_DEL_TIMER] running
on the
tree again if arrival interface for the group specified in a downstream router sends multicast join,
the timer is cancelled.

If a multicast JOIN_REQUEST through
it, or it is elected DR for one of its directly attached sub-
nets, received and subsequently receives an IGMP membership report.

4.3. Route Loops

Routing loops are only a concern when a router with at least one
child the QUIT_TIME bit (see
section 4.4.1) is attempting to re-join a CBT tree. In this case set on the re-
joining router sends a JOIN_REQUEST (subcode ACTIVE REJOIN) to arrival interface for the
best next-hop on specified
group, unset the path to an elected core. This join QUIT_TIME bit.

4.3. JOIN_ACK Processing

A JOIN_ACK is forwarded
as normal until it reaches either the specified core, another core,
or a on-tree router that mechanism by which an interface is already added to a
router's multicast forwarding cache; thus, the interface becomes part
of the group distribution tree. If the rejoin
reaches the primary core, loop detection

4.3.1. Sending JOIN_ACKs

The JOIN_ACK is not necessary because sent over the
primary never has a parent. same interface as the corresponding
JOIN_REQUEST was received. The primary core acks an active-rejoin by
means sending of a JOIN-ACK, subcode PRIMARY-REJOIN-ACK. This ack must be
processed by each router on the reverse-path of acknowledgement causes
the active-rejoin;
this ack creates tree state, just like a normal join-ack.

If an active-rejoin is terminated by any router on the tree other
than to add the primary core, loop detection must take place, as we now
describe.

If, in response interface to an active-rejoin, a JOIN-ACK its child interface list in its
forwarding cache for the group, if it is returned, subcode
NORMAL (as opposed to an ack with subcode PRIMARY-REJOIN-ACK), not already. If the router receiving the ack subsequently generates a JOIN-REQUEST, sub-
code NACTIVE-REJOIN (non-active rejoin). This packet serves only to
detect loops; it
does not create any transient yet have active state in the routers
it traverses, other than the originating for this group, this router (in case retransmis-
sions are necessary). Any on-tree must be
the core router receiving a non-active
rejoin is required to forward it over its parent interface for the
specified group. In this way, it will either reach the primary core,
which unicasts, directly to group; the sender, core creates a join ack with subcode PRI-
MARY-NACTIVE-ACK (so forwarding cache
entry and includes the sender knows no loop is present), or interface in its child interface list, and
sends the
sender receives the non-active rejoin it sent, via one of its child
interfaces, in which case JOIN_ACK downstream.

A JOIN_ACK is multicast or unicast, according to whether the rejoin obviously formed outgoing
interface supports multicast transmission or not.

4.3.2. Receiving JOIN_ACKs

The group and arrival interface must be matched to a loop. <group, ....,
upstream interface> from the router's cached transient state. If a loop no
match is present, the non-active join originator immediately
sends a QUIT_REQUEST to its newly-established parent and found, the loop JOIN_ACK is
broken.

Using figure 2 (over) to demonstrate this, if R3 discarded. If a match is attempting to re-
join found, a
CBT forwarding cache entry for the tree (R1 group is created, with "upstream
interface" marked as the core group's parent interface.

If "downstream interface" in figure 2) and R3 believes its best
next-hop to R1 the cached transient state is R6, and R6 believes R5 NULL, the
JOIN_ACK has reached the originator of the corresponding
JOIN_REQUEST; the JOIN_ACK is its best next-hop to R1,
which sees R4 as its best next-hop to R1 -- a loop not forwarded downstream. If "down-
stream interface" is formed. R3
begins by sending non-NULL, a JOIN_REQUEST (subcode ACTIVE_REJOIN, since R4 JOIN_ACK for the group is
its child) to R6. R6 forwards sent over
the join to R5. R5 "downstream interface" (multicast or unicast, accordingly). This
interface is on-tree for installed in the
group, so responds to child interface list of the active-rejoin with a JOIN-ACK, subcode NOR-
MAL (the ack traverses R6 on its way to R3).

R3 now generates a JOIN-REQUEST, subcode NACTIVE-REJOIN, and forwards
this group's
forwarding cache entry.

Once transient state has been confirmed by transferring it to its parent, R6. R6 forwards the non-active rejoin to R5, its
parent. R5 does similarly, as does R4. Now,
forwarding cache, the non-active rejoin has
reached R3, which originated it, so R3 concludes a loop transient state is present on deleted.

4.4. QUIT_NOTIFICATION Processing

A CBT tree is "pruned" in the parent direction downstream-to-upstream when-
ever a CBT router's child interface list for the specified group. It immediately sends a
QUIT_REQUEST group becomes NULL.

4.4.1. Sending QUIT_NOTIFICATIONs

A QUIT_NOTIFICATION is sent to R6, which in turn sends a quit if it has router's parent router on the tree
whenever the router's child interface list becomes NULL.

A QUIT_NOTIFICATION is not received
an ACK acknowledged; once sent, all information
pertaining to the group it represents is deleted from R5 already AND has itself the forwarding
cache after a child or subnets with member
presence. If so it does not send a quit -- the loop has been broken
by R3 sending the first quit.

QUIT_REQUESTs are typically acknowledged by means of short interval.

To ensure consistency between a QUIT_ACK. A child removes its and parent information immediately subsequent to send-
ing its first QUIT-REQUEST. The ack here serves to notify router given the (old)
child that it (the parent) has in fact removed its child information.
However,
potential for loss of a QUIT_NOTIFICATION, there might be cases where, due to failure, is a QUIT_TIME bit
associated with the parent can-
not respond. The child sends of each group entry; whenever a
QUIT_NOTIFICATION is sent for a QUIT-REQUEST group, the QUIT_TIME bit for that
group entry is set for a maximum of three
times, at PEND-QUIT-INTERVAL (5 sec) intervals.

------
| R1 |
------
|
---------------------------
|
------
| R2 |
------
|
---------------------------
| |
------ |
| R3 |--------------------------|
------ |
| |
--------------------------- |
| | ------
------ | | |
| R4 | |-------| R6 |
------ | |----|
| |
--------------------------- |
| |
------ |
| R5 |--------------------------|
------ |
|

Figure 2: Example Loop Topology

In another scenario [QUIT_TIME] seconds before the rejoin travels over a loop-free path,
entry is deleted and the
first on-tree router encountered QUIT_TIME bit unset. By default, this bit is
unset.

When the primary core, R1. In figure
2, R3 sends QUIT_TIME bit is set, if the router detects multicast traf-
fic for the group arriving over a join, subcode REJOIN_ACTIVE to R2, to-be-deleted parent interface (one
over which a quit has recently been sent), the next-hop on router sends another
QUIT_NOTIFICATION over that interface. This is multicast, or unicast,
as appropriate for the
path outgoing link. It continues to core R1. R2 forwards the re-join do so at
[QUIT_RATE] second intervals so long as data continues to R1, the primary core,
which returns arrive, and
provided [QUIT_TIME] has not yet expired.

If, after sending a QUIT_NOTIFICATION a JOIN-ACK, subcode PRIMARY-REJOIN-ACK, multicast JOIN_REQUEST for
the specified group arrives over the
reverse-path of interface the rejoin-active. Whenever a quit was sent, the
QUIT_TIME bit is immediately unset if it is set (any traffic arriving
over the interface will be for/from another child router receives attached to
the same link).

4.4.2. Receiving QUIT_NOTIFICATIONs

The group reported in the QUIT_NOTIFICATION must be matched with a PRI-
MARY-REJOIN-ACK
forwarding cache entry. If no loop detection match is necessary. found, the QUIT_NOTIFICATION
is ignored and discarded. If we assume R2 a match is on tree for found, if the corresponding group, R3 sends a
join, subcode REJOIN_ACTIVE to R2, which replies with a join ack,
subcode NORMAL. R3 must then generate a loop detection packet (join
request, subcode REJOIN-NACTIVE) which arrival inter-
face is forwarded to its parent,
R2, which does similarly. On receipt of the rejoin-Nactive, the pri-
mary core unicasts a join ack back directly to R3, with subcode PRI-
MARY-NACTIVE-ACK. This confirms to R3 that its rejoin does not form
a loop.

5. Data Packet Loops

The CBT protocol builds a loop-free distribution tree. If all routers
that comprise a particular tree function correctly, data packets
should never traverse a tree branch more than once (footnote 4).

CBT mode data packets from a non-member sender must arrive on a tree
via an "off-tree" interface. The CBT mode data packet's header
includes an "on-tree" field, which contains valid child interface in the value 0x00 until group entry, how the
data packet reaches an on-tree router. The first on-tree router must
convert this value to 0xff. This value remains unchanged, and from
here
proceeds depends on whether the packet should traverse only on-tree interfaces. QUIT_NOTIFICATION was multicast or
unicast.

If an
encapsulated packet happens to "wander" off-tree and back on again,
an on-tree router will receive the CBT encapsulated packet via an
off-tree interface. However, this router will recognise that QUIT_NOTIFICATION was unicast, the "on-
tree" field of corresponding child inter-
face is deleted from the encapsulating CBT header group's forwarding cache entry, and no fur-
ther processing is set to 0xff, required.

If the QUIT_NOTIFICATION was multicast, and so
immediately discards the packet.

_________________________
4 The exception to this is when CBT mode arrival interface is operating
between CBT routers connected to
a multi-access link; a
data packet may traverse the link in native mode (if
group members are present on the link), as well as CBT
mode valid child interface for sending the data between CBT routers on specified group, the
tree.

6. Data Packet Forwarding Rules

6.1. Native Mode

In native mode, when router sets a CBT
cache-deletion-timer [CACHE_DEL_TIMER].

Because this router receives might be acting as a data packet, parent router for multiple
downstream routers attached to the packet
may only be forwarded over outgoing tree interfaces (member subnets
and interfaces leading to outgoing on-tree neighbours) iff it has
been arrival link, [CACHE_DEL_TIMER]
interval gives those routers that did not send the
QUIT_NOTIFICATION, but received via a valid on-tree interface (or it over their parent interface, the packet has
arrived encapsulated
opportunity to ensure that the parent router does not remove the link
from its child interface list.

Therefore, on receipt of a non-member, i.e. off-tree, sender). Oth-
erwise, the packet is discarded.

Before multicast QUIT_NOTIFICATION over a packet is forwarded by parent
interface, a subnet's DR, provided the packet's
TTL is greater than 1, the packet's TTL is decremented.

6.2. CBT Mode

In CBT mode, routers ignore all non-locally originated native mode
multicast data packets. Locally-originated multicast data receiving router starts a random response interval timer
which is only
processed by set to [RND_RSP] seconds.

If a subnet's DR; in this case, the DR forwards the native multicast data packet, TTL 1, JOIN_REQUEST is received over any outgoing member subnets the same interface (par-
ent) for
which that router is DR. Additionally, the DR encapsulates same group before this router's [RND_RSP] timer expires,
it suppresses the
locally-originated multicasting of its own similar JOIN_REQUEST.

If a multicast and forwards it, CBT mode, over all tree
interfaces, as dictated by JOIN_REQUEST is not received via the CBT forwarding database.

When a router, operating in CBT mode, receives router's parent
link before [RND_RSP] expires, a CBT-mode encapsu-
lated data packet, it decapsulates one copy to send, native mode and
TTL 1, over any directly attached member subnets for which it is DR.
Additionally, an encapsulated copy JOIN_REQUEST is forwarded multicast over all outgoing
tree interfaces, as dictated by its CBT forwarding database.

Like the outer encapsulating IP header,
link for the previously quit group, with IP TTL value 1.

4.5. ECHO_REQUEST Processing

The ECHO_REQUEST message allows a child to monitor reachability to
its parent router for a group (or range of groups if the encapsu-
lating CBT header parent
router is decremented each time it the parent for multiple groups). Group information is processed by not
carried in ECHO_REQUEST messages.

4.5.1. Sending ECHO_REQUESTs

Whenever a router creates a CBT
router.

An example of CBT mode forwarding is provided towards cache entry due to the end receipt
of the
next section.

7. CBT Mode -- Encapsulation Details

In a multi-protocol environment, whose infrastructure may include
non-multicast-capable routers, it JOIN_ACK, the router begins the periodic sending of ECHO_REQUEST
messages over its parent interface. The ECHO_REQUEST is necessary multicast to tunnel data packets
between CBT-capable routers. This is called "CBT mode". Data packets
are de-capsulated by CBT routers (such that they become native mode
data packets) before being forwarded
the "all-cbt-routers" group over subnets with member hosts.
When multicasting (native mode) multicast-capable interfaces, and
unicast to member hosts, the TTL value of the
original IP header parent router otherwise.

ECHO_REQUEST messages are sent at [ECHO_INTERVAL] second intervals.
Whenever an ECHO_REQUEST is set to one. CBT mode encapsulation sent, [ECHO_INTERVAL] is as fol-
lows:

Figure 3. Encapsulation reset.

If, for CBT mode

The TTL value of the CBT header is set by the encapsulating CBT
router directly attached any echo-request sent to the origin of a data packet. This value parent, the expected response
(ECHO_REPLY) is decremented each time it not forthcoming within [ECHO_RTX_INTERVAL], the echo
request message is processed by a CBT router. An encap-
sulated data packet retransmitted. If no response is discarded when forthcoming
within [ECHO_TIMEOUT] seconds, the CBT header TTL value
reaches zero.

The purpose router sends a FLUSH_TREE message
over each of its child interfaces for the (outer) encapsulating IP header is to "tunnel"
data packets between CBT-capable routers (or "islands"). The outer IP
header's TTL value group, then removes all
forwarding cache state for the group.

4.5.2. Receiving ECHO_REQUESTs

If a ECHO_REQUEST is set to received over any valid child interface, the "length" of
receiving router responds with an ECHO_REPLY message over the corresponding tun-
nel, or MAX_TTL (255)if this same
interface. This message is not known, or subject multicast to change.

It is worth pointing out here the distinction between subnetworks and
tree branches (especially apparent in CBT mode), although they can be
one and the same. For example, a multi-access subnetwork containing
routers and end-systems could potentially be both a CBT tree branch "all-cbt-routers" group
over multicast-capable interfaces, and unicast otherwise.

If a subnetwork with group member presence. A tree branch which is
not simultaneously a subnetwork is either a "tunnel" or a point-to-
point link.

In CBT mode there are three forwarding methods used by CBT routers:

+o IP multicasting. This method sends an unaltered (unencapsulated)
data packet across a directly-connected subnetwork with group
member presence. Any host originating multicast data, does so
in this form.

+o CBT unicasting. This method is used ECHO_REQUEST message arrives via any valid parent
interface, the router resets its [ECHO_INTERVAL] timer for that
upstream interface, thereby suppressing the sending data packets
encapsulated (as illustrated above) across of its own
ECHO_REQUEST over that upstream interface.

4.6. ECHO_REPLY Processing

ECHO_REPLY messages allow a tunnel or point-to-
point link; child to monitor the IP destination address reachability of its
parent, and ensure the encapsulating IP
header group state information is a unicast address. En/de-capsulation takes place consistent between
them.

4.6.1. Sending ECHO_REPLY messages

An ECHO_REPLY message is sent in
CBT routers.

+o CBT multicasting. A CBT router on direct response to receiving an
ECHO_REQUEST message, provided the ECHO_REQUEST is received over any
one of this router's valid child interfaces. Additionally, an
ECHO_REPLY is sent periodically by a multi-access link can take
advantage parent router over each of multicast in its
child links, reporting all groups for which the case where multiple on-tree neigh-
bours link is its child.

ECHO_REPLY messages are reachable across unicast or multicast, as appropriate.

4.6.2. Receiving ECHO_REPLY messages

An ECHO_REPLY message must be received via a single physical link; valid parent interface.
When received, the outer
encapsulating IP header contains a multicast address as child router resets its des-
tination address. [ECHO_INTERVAL] timer for
this upstream interface. The IP module of end-systems on child router also caches the same link
subscribed to reported
"group report interval" (seconds) - the same time at which the next group
carrying ECHO_REPLY will discard these multicasts since
the CBT payload type (protocol id) of be sent by the outer IP header parent router. Like
[ECHO_INTERVAL], this is cached per upstream interface. If the group
carrying ECHO_REPLY does not
recognizable by hosts.

CBT routers create forwarding database (db) entries whenever they
send or receive arrive shortly after "group report
interval" has expired, a JOIN_ACK. The forwarding database describes the
parent-child relationships on a per-group basis. A forwarding
database entry dictates over which tree interfaces, and how (unicast
or multicast) a data packet is to be sent.

Note that a CBT forwarding db QUIT_NOTIFICATION is required sent for both CBT-mode and
native-mode multicasting.

Using our example topology in figure 1, let's assume each group for
which the CBT routers
are operating in CBT mode.

Member G originates an IP multicast (native mode) packet. R8 non-reporting router is the
DR for subnet S10. R8 therefore sends parent.

If this echo reply carries a (native mode, TTL 1) copy
over any member subnets list of groups, the child router must
match all those of its forwarding cache entries for which it is DR - S14 and S10 (the copy
over S10 the arrival
interface is the upstream interface. If the parent router does not sent, since
consider itself the packet was originally received from
S10). The multicast packet parent router for group(s) which the child thinks
is CBT mode encapsulated by R8, and uni-
cast to each of its children, R9 and R12; these children are not
reachable over parent, the same interface, otherwise R8 could have sent child sends a CBT
mode multicast. R9, the DR FLUSH_TREE message downstream for S12, need not IP multicast (native
mode) onto S12 since there are no
each such group. If this router has directly attached members present there. R9 unicasts for any
of the packet in CBT mode to R10, which is flushed groups, the DR receipt of an IGMP host membership report
for S13 and S15. R10
decapsulates the CBT mode packet and IP multicasts (native mode, TTL
1) to each any of S13 and S15.

Going upstream from R8, R8 CBT mode unicasts those groups will prompt this router to R4. It is DR for all
directly connected subnets and therefore IP multicasts (native mode) rejoin the data packet onto S5, S6 and S7, all of which have member pres-
ence. R4 unicasts, CBT mode, corre-
sponding tree(s).

If the upstream router considers itself the packet to all outgoing children, R3
and R7 (NOTE: R4 does not have a parent since it is for more groups
than does the primary core receiving router, this router sends a QUIT_NOTIFICATION
for each of those groups for which the group). R7 IP multicasts (native mode) onto S9. R3 CBT
mode unicasts to R1 and R2, its children. Finally, R1 IP multicasts
(native mode) onto S1 and S3, and R2 IP multicasts (native mode) onto
S4.

8. Non-Member Sending

For a multicast data packet to span beyond QUIT_TIME bit is set in the scope of
forwarding cache. Otherwise, the originat-
ing subnetwork at least one CBT-capable router must be present on
that subnetwork. takes no action.

4.7. FLUSH_TREE Processing

The DR for FLUSH_TREE (flush) message is the group on mechanism by which a router
invokes the subnetwork must encap-
sulate tearing down of all its downstream branches for a partic-
ular group. The flush message is multicast to the (native) IP-style packet "all-cbt-routers"
group when sent over multicast-capable interfaces, and unicast it to other-
wise.

4.7.1. Sending FLUSH_TREE messages

A FLUSH_TREE message is sent over each downstream (child) interface
when a core router has lost reachability with its parent router for the
group (footnote 5). The encapsulation required (detected via ECHO_REQUEST and ECHO_REPLY messages). All group
state is shown in figure 3;
CBT mode encapsulation removed from an interface over which a flush message is necessary so the receiving CBT router can
demultiplex the packet accordingly.

If
sent.

4.7.2. Receiving FLUSH_TREE messages

A FLUSH_TREE message must be received over the encapsulated packet hits parent interface for
the tree at an on-tree router, specified group, otherwise the
packet message is discarded.

The flush message must be forwarded according to the forwarding rules of section 6.1
or 6.2, depending on whether the receiving router is operating in
native- or CBT mode. Note that it is possible over each child interface for the different
interfaces of a router to operate in different (and independent)
modes.

If the first on-tree router encountered is
specified group.

Once the target core, various
scenarios define what happens next:

+o if flush message has been forwarded, all state for the target core group is not
removed from the primary, router's forwarding cache.

5. Timers and Default Values

This section provides a summary of the target core has
not yet joined the tree (because it has not yet itself received
any join-requests), timers described above,
together with their default values.

+o [HELLO_INTERVAL]: a base value making up the target core simply forwards bulk of the encapsu-
lated packet to the primary core; the primary core IP address is
included in the encapsulating CBT data packet header.

if the target core is not the primary, but has children, the
target core forwards the data according to the rules of section
6.
_________________________
5 It is assumed that CBT-capable routers discover
<core, group> mappings by means of some discovery pro-
tocol. Such inter-
val between sending a protocol is outside the scope of this
document. HELLO message. Default: 60 seconds.

+o if the target core is the primary, the primary forwards the data
according to the rules of section 6.2.

9. Eliminating the Topology-Discovery Protocol in the Presence of Tun-
nels

Traditionally, multicast protocols operating within a virtual topol-
ogy, i.e. an overlay [RND_RSP]: router's random response interval. Default: 2 sec-
onds.

+o [HELLO_TIMER]: (variable) interval between sending HELLO mes-
sages. [HELLO_TIMER] = [HELLO_INTERVAL + RND_RSP]

+o [HELLO_CONV]: convergence time of the physical topology, have required the
assistance one round of a multicast topology discovery protocol, such as that
present in DVMRP [1]. However, it is possible to have a multicast
protocol operate within a virtual topology without the need HELLO proto-
col. [HELLO_CONV] = [min(RND_RSP) + 2 seconds].

+o [JOIN_RTX_INTERVAL]: retransmission time for a
multicast topology discovery protocol. One way JOIN_REQUESTs.
Default: 5 seconds.

+o [JOIN_TIMEOUT]: time to achieve this is by
having a router configure all its tunnels raise exception due to its virtual neighbours
in advance. A tunnel is identified by a local interface address and a
remote tree join fail-
ure. Default: 3.5 times [JOIN_RTX_INTERVAL].

+o [CACHE_DEL_TIMER]: time to remove child interface address. Routing is replaced by "ranking" each such
tunnel from forward-
ing cache. Default: 2 seconds.

+o [QUIT_TIME]: time to remove parent interface associated with a particular core address; from forwarding
cache entry. Unset QUIT_TIME bit. Default: 60 seconds.

+o [QUIT_RATE]: period for sending QUIT_NOTIFICATION if the
highest-ranked route is unavailable (tunnel end-points are required
to run an Hello-like protocol traffic
persists. Default: 15 seconds.

+o [ECHO_INTERVAL]: interval between themselves) then the next-
highest ranked available route is selected, sending ECHO_REQUEST to parent
routers. Default: 60 seconds.

+o [ECHO_RTX_INTERVAL]: retransmission time for ECHO_REQUESTs.
Default 2 seconds.

+o [ECHO_TIMEOUT]: time to consider parent unreachable. Default:
3.5 times [ECHO_RTX_INTERVAL].

6. CBT Packet Formats and so on. The exact
specification of the Hello protocol is outside the scope of this doc-
ument. Message Types

CBT trees are built using the same join/join-ack mechanisms as
before, only now some branches of a delivery tree run in native mode,
whilst others (tunnels) run in CBT mode. Underlying unicast routing
dictates which interface a packet should be forwarded over. Each
interface is configured as either native mode or CBT mode, so a
packet can be encapsulated (decapsulated) accordingly.

As an example, router R's configuration would be as follows:

intf type mode remote addr
-----------------------------------
#1 phys native -
#2 tunnel cbt 128.16.8.117
#3 phys native -
#4 tunnel cbt 128.16.6.8
#5 tunnel cbt 128.96.41.1

core backup-intfs
--------------------
A #5, #2
B #3, #5
C #2, #4

The CBT forwarding database needs to be slightly modified to accommo-
date an extra field, "backup-intfs" (backup interfaces). The entry in
this field specifies a backup interface whenever a tunnel interface
specified in the forwarding db is down. Additional backups (should
the first-listed backup be down) are specified for each core in the
core backup table. For example, if interface (tunnel) #2 were down,
and the target core of a CBT control packet were core A, the core
backup table suggests using interface #5 as a replacement. If inter-
face #5 happened to be down also, then the same table recommends
interface #2 as a backup for core A.

10. CBT Packet Formats and Message Types

We distinguish between two types of CBT packet: CBT mode data pack-
ets, and CBT control packets. CBT control packets carry a CBT control
packet header.

CBT control packets control packets are encapsulated in IP, as illustrated below:

+++++++++++++++++++++++++++++++
| IP header | CBT control pkt |
+++++++++++++++++++++++++++++++

In CBT mode, the original data packet is encapsulated in a CBT header
and an IP header, as illustrated below:

The IP protocol field of the inner (original) IP header is used to
demultiplex a packet correctly; IP. CBT has been assigned IP
protocol number 7. The CBT module then demultiplexes based on the encapsulat-
ing CBT header's "type" field, thereby distinguishing between CBT
control packets and CBT mode data packets.

The CBT data packet header is illustrated below.

10.1. 7 by IANA [4].

6.1. CBT Common Control Packet Header Format (for

All CBT Mode data) control messages have a common fixed length header.

Figure 1. CBT Common Control Packet Header

This CBT specification is version 2.

CBT packet types are:

+o type 0: HELLO

+o type 1: JOIN_REQUEST

+o type 2: JOIN_ACK

+o type 3: QUIT_NOTIFICATION

+o type 4: ECHO_REQUEST

+o type 5: ECHO_REPLY

+o type 6: FLUSH_TREE

+o type 7: Bootstrap Message

+o type 8: Candidate Core Advertisement

+o Addr Length: address length in bytes of unicast or multicast
addresses carried in the control packet.

+o Checksum: the 16-bit one's complement of the one's complement
sum of the entire CBT control packet.

6.2. HELLO Packet Format

Figure 4. CBT Header

Each of the fields is described below: 2. HELLO Packet Format

HELLO Packet Field Definitions:

+o Vers: Version number -- this release specifies version 1. rnd response: random response interval in seconds.

+o preference: sender's HELLO preference.

+o option type: indicates CBT payload; values are defined for control
(0x00), and data (0xff). For the type of option present in the "option value"
field. One option type is currently defined: option type 0
(zero) = BR_HELLO; option value 0x00 (control), a CBT
control header 0 (zero); option length 0
(zero). This option type is assumed present rather than used with HELLO messages sent by a CBT header.
border router (BR) as part of designated BR election (see [5]).

+o hdr length: option len: length of the header, for purpose of checksum
calculation. "option value" field in bytes.

+o on-tree: indicates whether option value: variable length field carrying the packet is on-tree (0xff) or
off-tree (0x00). option value.

6.3. JOIN_REQUEST Packet Format

JOIN_REQUEST Field Definitions

+o checksum: the 16-bit one's complement of the one's complement group address: multicast group address of the CBT header, calculated across all fields.

+o IP TTL: TTL value corresponding to the value of the IP TTL
value of the original multicast packet, and set in the CBT
header by the DR directly attached to the origin host (decre-
mented by CBT routers visited).

+o group identifier: multicast group address.

+o first-hop router: identifies the encapsulating router
directly attached to the origin of a multicast packet. This
field is relevant to source-migration of being
joined. For a core to the source "wildcard" join (see Appendix A). It is set to NULL when core migration is
disabled.

+o primary core: the primary core for the group, as identified
by "group-id". This field is necessary for the case where
non-member senders happen to send to a secondary core, which
may not yet be joined to the primary core. This [5]), this field allows
the secondary to know which is the primary for the group, so
that the secondary can forward the (encapsulated) data
onwards to the primary.

+o T bit: indicates contains
the presence (1) or absence (0) of Type of
Service/flow-id value ("type", "length", "type of ser-
vice/flow-id") . INADDR_ANY.

+o S bit: indicates the presence (1) or absence (0) of a secu-
rity value ("type", "length", "security data").

10.2. Control Packet Header Format

The individual fields are described below. originating router: router that originated this JOIN_REQUEST.

Figure 5. CBT Control 3. JOIN_REQUEST Packet Header Format
+o Vers: Version number -- this release specifies version 1. target router: target (core) router for the group.

+o option type: indicates control message type (see sections 10.3).

+o code: indicates subcode allows the specification of control message type.

+o # cores: number a variety of core addresses carried by this control
packet.

+o header length:
JOIN_REQUEST options. One option is currently defined: option
type 0 (zero) = BR_JOIN; option length of the header, 0 (zero); option value 0
(zero). This option is used by a CBT domain border router to
join an internal core for purpose of checksum
calculation. all groups that map to that core. The
state instantiated by a JOIN_REQUEST with this option set is
represents (*, core). For further details, see [5].

6.4. JOIN_ACK Packet Format

0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CBT Control Packet Header |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| group address |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| target router |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| option type | option len | option value |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Figure 4. JOIN_ACK Packet Format
JOIN_ACK Field Definitions

+o checksum: the 16-bit one's complement group address: multicast group address of the one's complement
of group being
joined.

+o target router: router (DR) that originated the corresponding
JOIN_REQUEST.

6.5. QUIT_NOTIFICATION Packet Format

0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CBT control header, calculated across all fields. Control Packet Header |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| group address |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| originating child router |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Figure 5. QUIT_NOTIFICATION Packet Format

QUIT_NOTIFICATION Field Definitions

+o group identifier: address: multicast group address.

+o address of the group mask: mask value for aggregated CBT joins/join-acks.
Zero for non-aggregated joins/join-acks. being
joined.

+o packet origin: originating child router: address of the CBT router that originated originates
the
control packet. QUIT_NOTIFICATION.

6.6. ECHO_REQUEST Packet Format

ECHO_REQUEST Field Definitions

+o primary core address: the originating child router: address of the primary core for router that originates
the
group.

+o target core address: desired core affiliation of control mes-
sage.

+o Core #N: IP address for each of a group's cores.

+o T bit: indicates the presence (1) or absence (0) of Type of
Service/flow-id value ("type", "length", "type of ser-
vice/flow-id") .

+o S bit: indicates the presence (1) or absence (0) of a secu-
rity value ("type", "length", "security data").

10.3. ECHO_REQUEST.

0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CBT Control Message Types

There are ten types of CBT message. All are encoded in the CBT con-
trol header, shown in figure 5.

+o JOIN-REQUEST (type 1): generated by a router and unicast to
the specified core address. It is processed hop-by-hop on its
way to the specified core. Its purpose is to establish the Packet Header |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| originating child router |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Figure 6. ECHO_REQUEST Packet Format
6.7. ECHO_REPLY Packet Format

0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CBT router, and all intermediate CBT routers, as
part of the corresponding delivery tree. Note that all cores
for the corresponding Control Packet Header |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| originating parent router |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| group are carried in join-requests. report interval | num groups |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| group address #1 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| group address #2 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ...... |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| group address #n |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Figure 7. ECHO_REPLY Packet Format

ECHO_REPLY Field Definitions

+o JOIN-ACK (type 2): an acknowledgement to the above. The full
list oringinating parent router: address of core addresses is carried in a JOIN-ACK, together
with the actual core affiliation (the join may have been ter-
minated by an on-tree router on its journey to the specified
core, and the terminating router may or may not be affiliated
to the core specified in the original join). A JOIN-ACK tra-
verses the reverse path as the corresponding JOIN-REQUEST,
with each CBT router on the path processing the ack. It is
the receipt of a JOIN-ACK that actually "fixes" tree state.

+o JOIN-NACK (type 3): a negative acknowledgement, indicating
that the tree join process has not been successful.

+o QUIT-REQUEST (type 4): a request, sent from a child to a par-
ent, to be removed as a child of that parent.

+o QUIT-ACK (type 5): acknowledgement to the above. If the par-
ent, or the path to it is down, no acknowledgement will be
received within the timeout period. This results in the
child nevertheless removing its parent information.

+o FLUSH-TREE (type 6): a message sent from parent to all chil-
dren, which traverses a complete branch. This message results
in all tree interface information being removed from each
router on the branch, possibly because of a re-configuration
scenario.

+o CBT-ECHO-REQUEST (type 7): once a tree branch is established,
this messsage acts as a "keepalive", and is unicast from
child to parent (can be aggregated from one per group to one
per link. See section 4).

+o CBT-ECHO-REPLY (type 8): positive reply to the above.

+o CBT-BR-KEEPALIVE (type 9): applicable to border routers only.
See [11] for more information.

+o CBT-BR-KEEPALIVE-ACK (type 10): acknowledgement to the above.

10.3.1. CBT Control Message Subcodes

The JOIN-REQUEST has three valid subcodes:

+o ACTIVE-JOIN (code 0) - sent from a CBT router that has no
children for the specified group.

+o REJOIN-ACTIVE (code 1) - sent from a CBT router that has at
least one child for the specified group.

+o REJOIN-NACTIVE (code 2) - generated by a router subsequent to
receiving a join ack, subcode NORMAL, in response to a
active-rejoin.

A JOIN-ACK has three valid subcodes:

+o NORMAL (code 0) - sent by a core router, or on-tree router,
acknowledging joins with subcodes ACTIVE-JOIN and REJOIN-
ACTIVE.

+o PRIMARY-REJOIN-ACK (code 1) - sent by a primary core to
acknowledge the receipt of a join-request received with sub-
code REJOIN-ACTIVE. This message traverses the reverse-path
of the corresponding re-join, and is processed by each router
on that path.

+o PRIMARY-NACTIVE-ACK (code 2) - sent by a primary core to
acknowledge the receipt of a join-request received with sub-
code REJOIN-NACTIVE. This ack is unicast directly to the
router that generated the rejoin-Nactive, i.e. the ack it is
not processed hop-by-hop.

11. CBT Protocol Number

CBT has been assigned IP protocol number 7. CBT control messages are
carried directly over IP.

12. Default Timer Values

There are several CBT control messages which are transmitted at fixed
intervals. These values, retransmission times, and timeout values,
are given below. Note these are recommended default values only, and
are configurable with each implementation (all times are in seconds):

+o CBT-ECHO-INTERVAL 30 (time between sending successive CBT-ECHO-
REQUESTs to parent).

+o PEND-JOIN-INTERVAL 5 (retransmission time for join-request if no
ack rec'd)

+o PEND-JOIN-TIMEOUT 30 (time to try joining a different core, or
give up)

+o EXPIRE-PENDING-JOIN 90 (remove transient state for join that has
not been ack'd)

+o PEND_QUIT_INTERVAL 5 (retransmission time for quit-request if no
ack rec'd)

+o CBT-ECHO-TIMEOUT 90 (time to consider parent unreachable)

+o CHILD-ASSERT-INTERVAL 90 (increment child timeout if no ECHO
rec'd from a child)

+o CHILD-ASSERT-EXPIRE-TIME 180 (time to consider child gone)

+o IFF-SCAN-INTERVAL 300 (scan all interfaces for group presence.
If none, send QUIT)

+o BR-KEEPALIVE-INTERVAL 200 (backup designated BR to designated BR
keepalive interval)

+o BR-KEEPALIVE-RETRY-INTERVAL 30 (keepalive interval if BR fails
to respond)

13. Interoperability Issues

Interoperability between CBT and DVMRP has recently been defined in
[11].

Interoperability with other multicast protocols will be fully speci-
fied as the need arises.

14. CBT Security Architecture

see [4].

Acknowledgements

Special thanks goes to Paul Francis, NTT Japan, for the original
brainstorming sessions that brought about this work.

Thanks too to Sue Thompson (Bellcore). Her detailed reviews led to
the identification of some subtle protocol flaws, and she suggested
several simplifications.

Thanks also to the networking team at Bay Networks for their comments
and suggestions, in particular Steve Ostrowski for his suggestion of
using "native mode" as a router optimization, and Eric Crawley.

Thanks also to Ken Carlberg (SAIC) for reviewing the text, and gener-
ally providing constructive comments throughout.

I would also like to thank the participants of the IETF IDMR working
group meetings for their general constructive comments and sugges-
tions since the inception of CBT.

APPENDICES

DISCLAIMER: As of writing, the mechanisms described in Appendices A and
B have not been tested, simulated, or demonstrated.

APPENDIX A

Dynamic Source-Migration of Cores

A.0 Abstract

This appendix describes CBT protocol mechanisms that allow a CBT mul-
ticast tree, initially constructed around a randomly-placed set of
core router, to dynamically reconfigure itself in response to an
active source, such that the CBT tree becomes rooted at the source's
local CBT router. Henceforth, CBT emulates a shortest-path tree.

For clarity, the mechanisms are described in the context of "flat"
multicasting, but are transferrable to a hierarchical model with only
minor changes.

A.1 Motivation

One of the criticisms levelled against shared tree multicast schemes
is that they potentially result in sub-optimal routes between
receivers. Another criticism is that shared trees incur a high traf-
fic concentration effect on the core routers. Given that any shared
tree is likely to have two, three, or more cores which can be strate-
gically placed in the network, as well as the fact that any on-tree
router can act as a "branch point" (or "exploder point"), shared tree
traffic concentration can be significantly reduced. This note never-
theless addresses both of these criticisms by describing new mecha-
nisms that

+o allow a CBT to dynamically transition from a random configura-
tion to one where any CBT router can become a core - more pre-
cisely, that which is local to a source, and...

+o remove the traffic concentration issue completely, as a result
of the above; traffic concentration is not an issue with source-
rooted trees.

The mechanisms described here are relevant to non-concurrent sources;
the concurrent-sender case is not addressed here, although experience
with MBONE applications for the past several years suggests that most
multicast applications are of the single, infrequently-changing
sender type. Also, it is not necessarily implied that the initial
CBT tree must be transitioned. Any transition is an "all-or-nothing"
transition, meaning that either all the tree transitions, or none of
it does (footnote 6).

A.2 Goals & Requirements

By means of the mechanisms described, this Appendix sets out to
achieve the follwoing:

+o provide mechanisms that allow the dynamic transition from an
initial CBT, constructed around a pre-configured set of cores,
to a CBT that is rooted at a core attached to a sender's local
subnetwork. This is source-rooted tree emulation.

+o ensure that these mechanisms do not impact CBT's simplicity or
scalability.

+o eliminate completely the traffic concentration issue from CBT.

+o to eliminate the core placement/core advertisement problems.

+o ensure that the scheme is robust, such that if a source's local
router (or link to it) should fail, the CBT self-organises
itself and returns to its original configuration.

+o the mechanisms should provide the same even to non-member
senders.

The above incurs a few additional requirements on existing baseline
CBT mechanisms described in this specification:

+o a new JOIN-REQUEST subcode, REVERSE-JOIN

+o a new JOIN-ACK subcode, REVERSE-ACK
_________________________
6 This is the expected behaviour of PIM Sparse Mode;
on reciept of high-bandwidth traffic, most receivers'
local routers will be configured to transition to
source trees.

+o new JOIN-ACK subcode, CORE-MIGRATE

+o a "first-hop router" field needs to be included in the CBT data
packet header.

+o a new message type:

- SOURCE-NOTIFICATION

+o CBT-mode data encapsulation is required until the local CBT
router connected to an active source receives a JOIN-REQUEST,
whose "target core address" field is one of its own IP
addresses.

These new additions are explained in the next section.

A.3 Source-Tree Emulation Criteria

CBT routers are configured with a lower-bound data-rate threshold
that is the expected boundary between low- and high-bandwidth data
rate traffic. CBT also monitors the duration each sender sends. If
this duration exceeds a pre-configured value (global across CBT), say
3 minutes, AND the data rate threshold is exceeded, the CBT tree
transitions such that receivers become joined to the "core" local to
the source's subnet, i.e. the CBT tree becomes source-rooted, but
nevertheless remains a CBT.

A.4 Source-Migration Mechanisms
E o o D
\ /
\ /
L o \ /
\ o C
\ N /
\ /
\A(2) (1)B /
O===================================O
| |
M | |
| |
K o o H
/\ /\
/ \ / \
/ \ / \
s J o o I G o o F
----------

Key: B = primary core
A = secondary core
s = sending host
J = sending host's local DR
M & N = network nodes not on original CBT tree

Figure A1: Original CBT Tree

In figure A1, host s starts sending native mode multicast data. CBT
router J encapsulates it as CBT mode, inserting its own IP address in
the "first-hop router" field of the CBT mode data packet header. This
data packet flows over the CBT tree.

Note that tree migration can be disabled either by sending all pack-
ets in native mode, or by inserting NULL value into the "first-hop
router" field. Since the first-hop router is the original encapsulat-
ing router (data packets are always originated from hosts in native
mode), the first-hop router knows whether the sender's data rate war-
rants activating the "first-hop router" field; for the purpose of the
ensuing protocol description, we assume this is the case.

Any router on the tree receiving the CBT mode data packet, inspects
the "first-hop router" field of the CBT header, and compiles a join-
request to send to it. In order to fully specify the join, it must
inspect its underlying unicast routing table(s) to find the best
next-hop to the source's first hop router. That next hop will be
either on or off the existing CBT tree for the group. If the next hop
is off-tree, the join generated is given a subcode of ACTIVE-JOIN (as
per CBT spec), and a "target core address" of the source's first hop
router. The join is then forwarded and processed according to the CBT
specification. The primary core, and the original core list, remain
specified in their respective fields of the CBT control packet
header.

Using figure A1 to illustrate an example, node L's routing tables
suggest that the best next-hop to J, the source's first hop router,
is via node M, not yet on the tree. So, node L generates a join and
forwards it to M, which forwards it to J. The join-ack (subcode NOR-
MAL) returns to L via M on the reverse-path of the join. When the
join-ack reaches L, L sends a QUIT-REQUEST to A, its old parent. The
shortest-path branch now exists, L-M-J.

If the best next hop to the source's first hop router is via an
existing on-tree interface, if that interface is the node's parent on
the current tree, no further action need be taken, and no join need
be sent towards the source, J.

However, the join's best next hop may be via an existing child inter-
face - this is where the new join type, subcode REVERSE-JOIN, comes
in. The purpose of this join type is to simply reverse the existing
parent-child relationship between two adjacent on-tree routers; each
end of the link between the two routers is re-labelled. This join
must be acknowledged by means of a JOIN-ACK, subcode REVERSE-ACK. A
reverse-join is only ever sent from a child to its parent.

Immediately subsequent to sending a reverse-join-ACK, the sending
node's old parent interface is labelled as "pending child", and a
timer is set on that interface. This is a delay timer, set at a
default of 5 seconds, during which time a reverse-join is expected
over that interface from the node's old parent. Should this timer
expire, a REVERSE-ASSERT message is sent to the old parent (new
child) to cause it to agree to the change in the parent-child rela-
tionship. A REVERSE-ASSERT must be ack'd (REVERSE-ASSERT-ACK). If,
after (say) three retransmissions (at 5 sec intervals) no reverse-
assert-ack has been received, a QUIT-REQUEST is sent to the old par-
ent and the corresponding interface is removed from this node's cur-
rent forwarding database.

Of course, if a node has already received a reverse-join during the
period one of its other interfaces was changing its parent-child
relationship with another of its neighbours, then the pending-child
delay timer need not be activated.

Looking at figure A1 again, here's the process of how the parent-
child relationships change on the tree when an active source, s,
starts sending. Of course, links E-C, I-J, and L-J do not do this
because they forge completely new paths towards the source's local
router, J.

K sends a reverse-join to J. J acks this with a join-ack, subcode
REVERSE-ACK. At this point, J is K's parent, and I is still K's
child. K now sets the pending-child delay timer on its interface to
A (K's old parent), and expects a reverse-join from A. If it weren't
to arrive after the delay timer expires, plus several retransmissions
of a reverse-assert control message, K can send a quit to A (it sends
a quit because, as far as A is concerned, it thinks K is still its
child) and removes the K-A interface from its CBT forwarding
database. However, assuming a reverse-join does arrive at K from A
before the delay timer expires, K acks the reverse-join and cancels
the delay timer on that interface.

Next, let's consider CBT router (node) I. I's unicast routing table
suggest it can reach J directly (next-hop) via a different interface
than the I-K interface, so I sends a join-request, subcode active-
join, to J, which acks it as normal. On receipt of the ack, I sends a
quit to K and removes K as its parent from its database.

Now let's consider node L. Like I, it finds a new path to J, via M,
so simply sends a new join to J, via M, and on receipt of the join-
ack, sends a quit to A, and removes A from its forwarding database.
A new, shortest-path, branch now exists, J-M-L.

Next let's consider A-B, the link between the cores. A is the sec-
ondary, and B is the primary, so A originally joined towards B. So,
B sends a reverse-join to A. A sends a reverse-ack to B, so A is now
B's parent, and B has children B-H, and B-C. Note that the role of
primary and secondary is not affected - the target of B's join to A
is the source's local router, J.

The existing branches D-C-B, F-H-B, and G-H-B, need not change any of
their parent-child relationships, since each of these nodes' unicast
routing tables indicate that the best next-hop a join-request, tar-
getted at source J, would take, is via the corresponding existing
parent.

For E, it sends a new join via N to J. On receipt of the join-ack, it
sends a quit to C. A new branch has been created, E-N-J.

Each node on the tree now has a shortest-path to J, the source's
local CBT router. Hence, J is the root ("core") of a shortest-path
multicast tree.

Note that these new mechanisms augment the CBT protocol, and the
baseline CBT protocol engine is not affected in any way by this add-
on mechanism.

A.5 Robustness Issues

Some immediate questions might be:

+o what happens to the source-rooted tree if the source's local CBT
router fails?

+o what happens if the source's local CBT router fails whilst the
initial tree is transitioning?

+o what happens if the tree is partitioned, or not yet fully con-
nected, when a source starts sending?

+o how do new receivers join an already-transitioned tree?

All of these questions are now addressed:

+o What happens to the source-rooted tree if the source's local CBT
router fails?

A source-rooted CBT has a single point of failure - the root of
the tree.

In spite of a source being joined, the corelist (primary & sec-
ondaries) is carried in CBT control packets, as per the CBT
spec. However, the contents of the "target core address" field
identifies the IP address of the source's local CBT router. So,
in the event of a failure, the CBT routers still have all the
information they need to rejoin the original tree, constructed
around the corelist. Rejoining then, proceeds according to the
rules of the CBT specification.

Of course, rejoining the original tree happens only after sev-
eral attempts have been made to rejoin the source's "core".

+o What happens if the source's local CBT router fails whilst the
initial tree is transitioning?

This really is no different to the above case. The parts of the
tree that have transitioned will rejoin the original tree
according to their corresponding corelist. Those parts of the
tree in the process of transitioning may temporarily transition,
but eventually those nodes will receive a FLUSH from a CBT
router adjacent to the failed source router ("core"). They then
rejoin the original tree.

+o What happens if the tree is partitioned, or not yet fully con-
nected, when a source starts sending?

The problem here is that some parts of the network (CBT tree)
may not receive CBT encapsulated mode data packets before the
source's local DR starts forwarding data in native mode, and so
those receivers will not know the IP address of the local DR to
join to.

For example, assume a secondary core with downstream members
cannot reach the primary. If the routers adjacent to the secon-
daries are all functioning correctly, the secondaries themselves
may not be aware that a partition has occurred somewhere further
upstream. So, what if a source downstream from a secondary,
starts sending data after the partition has happened?

A new control message, the SOURCE-NOTIFICATION, is used to solve
this problem. As soon as any core recieves CBT mode encapsulated
data, it caches the source "core" IP address, and starts multi-
casting (to the group) SOURCE-NOTIFICATION messages, one every
minute. Source-notifications contain the IP address of the
source's local DR. A core continues to multicast source-
notications at 1 minute intervals until the source has ceased
transmitting data for more than 20 seconds.

Obviously, if a CBT is fully connected, the larger proportion of
source-notifications will be redundant. However, this cost jus-
tifies the robustness the scheme provides.

If an off-tree source begins sending data, which first hits the
tree at a secondary core with no receivers attached, the
secondary does not trigger a join towards the primary, but
instead just unicasts the data, in CBT mode, to the primary (as
per CBT spec). The primary then forwards the data over any con-
nected tree branches. Receivers can then begin transitioning. In
this way, a transitioned CBT tree extends to the first hop
router of a non-member sender.

Note that cores and on-tree routers only ever react to active
sources iff they have an existing CBT forwarding database for
the said group. For example, a primary core would not establish
a shortest-path branch to a non-member sender unless it has at
least one existing child registered for the corresponding group.

+o How do new receivers join an already-transitioned CBT?

New receivers will always attempt to join one of the cores in
the corelist for a group. Two things can happen here: firstly, a
new join, targetted at one of the cores in the corelist eventu-
ally reaches that target core. Secondly, the new join hits a
router already established on-tree, but the router encountered
is now joined to the source tree (source "core").

For the first scenario, all on-tree routers and all core routers
maintain the address of which upstream core their CBT branch
actually emanates from (as per CBT spec). When a new join
arrives at one of the original cores, the core checks whether
its own current core affiliation is to a core outside the
corelist set. If so, that core is a source "core", so the core
responds to the new join with a JOIN-ACK, subcode CORE-MIGRATE.
This join-ack contains the address of the active source "core".
This join-ack causes a join-request to be issued by one of the
routers that receives it - the router whose path to the core
(just joined) diverges from that to the source "core"; this can
easily be gleaned from unicast routing. The router then simply
directs it new join at the source "core", and on receipt of the
join-ack, sends a quit to its now "old" parent.

For the second case, the solution is trivial; any on-tree router
receiving a join targetted either at one of the original cores
for the group, or the active source "core", simply acks (subcode
NORMAL) the join and includes in the ack the source "core"
affiliation (as per CBT spec).

A.6 Loops
It may seem that the potential for a transitioning tree to form
loops, especially in the presence of reverse-joins, is greatly
increased. This is probably NOT the case; "reversed branches" are
those that are already part of a loop-free tree that CBT constructs
around the original set of cores. Transitioned tree are just CBTs,
whereby the core is simply rooted at the source. Loops are no more
likely with these mechanisms then they are with baseline CBT. Note
that these are assertions - formal proofs may be more appropriate.

APPENDIX B

Group State Aggregation

B.1 Introduction

Although the scalability of shared tree multicast schemes is attrac-
tive now, to scale over the longer-term, a combination of hierarchy
(support mechanisms that facilitate domain-oriented multicasting),
and group aggregation strategies, is required. If IP multicast is to
have a long-term future in the Internet as a global transport mecha-
nism, by far the most serious challenge is to address the issue of
group state aggregation.

Shared trees were developed partly to address scalability with
regards to multicast state maintained in the network, which resulted
in an improvement in that state by a factor of the number of active
sources (a source being a subnetwork aggregate). However, it is per-
ceived that the number of sources sending to any one group will not
grow as fast as the number of groups, indeed the latter will probably
grow at several orders of magnitude faster [12]. Therefore, it is
essential to contain this potential problem, particularly for the
benefit of routers on wide-area links, by designing an effective
group state aggregation mechanism, capable of collapsing group state.

Unlike unicast addresses, multicast addresses cannot be aggregated
according to topological locality; multicast addresses are truly
location-independent. Thus, it would not seem obvious how the problem
can be addressed - clearly, it must be looked at in a different way.

In order to be effective, flexibility and efficiency must be facets
of group aggregation; an aggregation scheme must be able to accommo-
date groups with wide-ranging characteristics in the least constrain-
ing way possible. For example, the trend towards small, non-local
groups (e.g. 4 or 5 person audio/video conferences between different
user groups spread over different countries/continents); it is these
types of groups that are likely to result in an explosive growth in
state. Also, these groups will, in all likelihood, utilize multicast
addresses that are randomly spread across the multicast address
space, making aggregation seemingly more difficult. An aggregation
scheme must therefore account for this.

B.2 Design Overview
This scheme involves replacing a subset the router originating
this ECHO_REPLY.

+o group report interval: number of individual tree state pre-
sent on inter-domain links, and aggregating it over seconds until the sending
router will send its next ECHO_REPLY containing a single shared
tree. The scheme does not yet specify how candidate list of group
addresses.

+o num groups: the number of groups for aggre-
gation are arrived at, but an obvious scheme to would be to aggregate
already-overlapping distribution trees. The pivotal idea behind being reported by this
approach encompasses two inter-dependent strategies:
ECHO_REPLY.

+o administratively defining group address: a portion list of the multicast address
space group addresses for aggregate groups. For brevity, an example might be which
this router considers itself a parent router w.r.t. the
range 238.0.0.0 - 238.255.255.255. link
over which this message is sent.

6.8. FLUSH_TREE Packet Format

Figure 8. FLUSH_TREE Packet Format

FLUSH_TREE Field Definitions

+o associated with each aggregate group address: multicast group address is a mask, specify-
ing the portion of the address that it used to identify the
aggregate group itself (the portion covered by being
"flushed".

7. Core Router Discovery

For intra-domain core discovery, CBT has decided to adopt the mask); "boot-
strap" mechanism currently specified with the
remaining address space PIM sparse mode proto-
col [2]. This bootstrap mechanism is used scalable, robust, and does not
rely on underlying multicast routing support to deliver core router
information; this information is distributed via traditional unicast
hop-by-hop forwarding.

It is expected that the bootstrap mechanism will be specified inde-
pendently as an index a "generic" RP/Core discovery mechanism in its own sepa-
rate document. It is unlikely at this stage that the bootstrap mecha-
nism will be appended to an ordered list a well-known network layer protocol, such as
IGMP [3], though this would facilitate its ubiquitous (intra-domain)
deployment. Therefore, each multicast routing protocol requiring the
bootstrap mechanism must implement it as part of the multicast rout-
ing protocol itself.

A summary of groups with which the aggregate address is associated. The
ordered list and its association with a group aggregate address
is conveyed by means operation of a protocol message (TBD). The index the bootstrap mechanism follows
(details are provided in [7]). It is
used to de-aggregate at region boundaries (border routers).

The scheme subscribes to assumed that all routers within
the domain implement the notion "bootstrap" protocol, or at least forward
bootstrap protocol messages.

A subset of aggregation-on-demand; a bor-
der router (BR) is the domain's routers are configured with a threshold number of groups on a
BRs external interface, above which it begins to solicit aggregations
periodically, say once be CBT candidate
core routers. Each candidate core router periodically (default every hour.

As an example, say BR 123 wishes
60 secs) advertises itself to aggregate 200 groups. BR 123 ran-
domly chooses (or by some address allocation algorithm) a group
aggregate address. It has been established that the number of groups
for which aggregation is desired is 200. domain's Bootstrap Router (BSR),
using "Core Advertisement" messages. The nearest power of 2 value
to 200 BSR is 256 (2^8), and so itself elected
dynamically from all (or participating) routers in the aggregate mask covers 24 bits, leav-
ing 8 domain. The
domain's elected BSR collects "Core Advertisement" messages from can-
didate core routers and periodically advertises a candidate core set
(CC-set) to specify each individual group's traffic flowing over the
aggregate tree.

So we have:
Group aggregate address: 238.10.12.0

Group aggregate mask: 238.10.12/24

A data packet for the 30th listed group (listed other router in a protocol message
(TBD) as described above) would be addressed to: 238.10.12.30.

Similarly, a data packet pertaining to the 150th listed group would
be addressed to: 238.10.12.150, and so on.

All routers comprising the aggregate tree need only maintain domain, using traditional hop-
by-hop unicast forwarding. The BSR uses "Bootstrap Messages" to
advertise the
group aggregate address CC-set. Together, "Core Advertisements" and mask, together with "Bootstrap
Messages" comprise the aggregate tree's
associated interfaces. If "bootstrap" protocol.

When a number of individual shared trees have
been replaced by router receives an aggregate tree, then the core routers (RPs) of
each IGMP host membership report from one of those shared trees must additionally maintain its
directly attached hosts, the complete
list local router uses a hash function on the
reported group address, the result of groups associated with an <aggregate address/mask-len> so which is used as an index into
the CC-set. This is how local routers discover which core to be able to "re-direct" any incoming joins use for already aggregated
groups. Similarly, border routers (BRs) are incurred
a particular group.

Note the storage
cost hash function is specifically tailored such that a small
number of maintaining the individual consecutive groups associated with an <aggre-
gate address/mask-len>, so as to be able always hash to aggregate and de-
aggregate as data packets flow across a (sub)region's border.

B.3 Scaling Further

The scheme described the same core. Further-
more, bootstrap messages can be applied recursively (to border routers) carry a "group mask", potentially limit-
ing a CC-set to accommodate a hierarchy containing an arbitrary number particular range of levels.

The scheme described imposes two general requirements (or assump-
tions):

+o groups. This can help reduce
traffic concentration at the core.

If a well defined aggregate group address space BSR detects a particular core as being unreachable (it has not
announced its availability within some period), it deletes the rele-
vant core from the CC-set sent in its next bootstrap message. This is
how a local router discovers a group's core is unreachable; the
router must re-hash for each level of
hierarchy (or scope levels).

+o affected group and join the ability to arbitrarily create boundaries in multicast
routers, thereby separating different hierarchical levels. new core
after removing the old state. The former will require consensus within removal of the IETF "old" state follows
the sending of a QUIT_NOTIFICATION upstream, and a FLUSH_TREE message
downstream.

7.1. Bootstrap Message Format

0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CBT common control packet header |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| For full Bootstrap Message specification, see [7] |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Figure 9. Bootstrap Message Format

7.2. Candidate Core Advertisement Message Format

0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CBT common control packet header |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| For full Candidate Core Adv. Message specification, see [7] |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+

Figure 10. Candidate Core Advertisement Message Format

8. Interoperability Issues

Interoperability between CBT and approval from
the IANA. The latter capability DVMRP is already available in multicast
routers; boundaries are specified in a multicast routers configura-
tion file. This capability is currently available in the best known
multicast routing protocols: DVMRP, M-OSPF, PIM, and CBT.

Defining boundaries may require some degree of coordination; whenever
a particular scoped level (boundary) is introduced which has multiple
entry/exit [5].

Interoperability with other multicast routers, these must all be configured such that
their boundary definitions are identical, i.e. they must each protocols will be con-
figured with fully speci-
fied as the same boundary-address/mask (the range 239.0.0.0 -
239.255.255.255 is need arises.

Acknowledgements

Special thanks goes to Paul Francis, NTT Japan, for the IANA-defined multicast boundary address
range).

Author Information:

Tony Ballardie,
Department original
brainstorming sessions that brought about this work.

Others that have contributed to the progress of Computer Science,
University College London,
Gower Street,
London, WC1E 6BT,
ENGLAND, U.K.

Tel: ++44 (0)71 419 3462
e-mail: A.Ballardie@cs.ucl.ac.uk

Scott Reeve, CBT include Ken Carl-
berg, Eric Crawley, Nitin Jain,
Bay Networks, Inc.
3, Federal Street,
Billerica, MA 01821,
USA.

Tel: ++1 508 670 8888
e-mail: {sreeve, njain}@BayNetworks.com Steven Ostrowsksi, Radia Perlman,
Scott Reeve, Clay Shields, Sue Thompson, Paul White.

The participants of the IETF IDMR working group have provided useful
feedback since the inception of CBT.

References

[1] T. Pusateri. Distance Vector Core Based Trees (CBT) Multicast Routing Protocol. Architecture;
A. Ballardie; ftp://ds.internic.net/internet-drafts/draft-ietf-idmr-
cbt-arch-**.txt. Working draft, June 1996. (draft-ietf-idmr-dvmrp-v3-01.{ps,txt}). 1997.

[2] J. Moy. Multicast Routing Extensions to OSPF. Communications of
the ACM, 37(8): 61-66, August 1994. Also RFC 1584, March 1994.

[3] D. Farinacci, S. Deering, D. Estrin, and V. Jacobson. Protocol Independent Multicast (PIM) Dense-Mode Specification. Working draft,
July 1996. (draft-ietf-idmr-pim-dm-spec-02.{ps,txt}).

[4a] A. Ballardie. Core Based Tree (CBT) Multicast Architecture. Sparse Mode/Dense Mode; D.
Estrin et al; ftp://netweb.usc.edu/pim Working draft, July drafts, 1996. (draft-ietf-idmr-cbt-arch-04.txt)

[4] A. J. Ballardie. Scalable Multicast Key Distribution; RFC 1949,
SRI Network Information Center, 1996.

[5] A. J. Ballardie. "A New Approach to Multicast Communication in a
Datagram Internetwork", PhD Thesis, 1995. Available via anonymous ftp
from: cs.ucl.ac.uk:darpa/IDMR/ballardie-thesis.ps.Z.

[6] W. Fenner.

[3] Internet Group Management Protocol, version 2 (IGMPv2). (IGMPv2); W. Fenner;
ftp://ds.internic.net/internet-drafts/draft-ietf-idmr-igmp-v2-**.txt.
Working draft, May 1996. (draft-idmr-igmp-v2-03.txt).

[7] B. Cain, S. Deering, A. Thyagarajan. Internet Group Management
Protocol Version 3 (IGMPv3) (draft-cain-igmp-00.txt).

[8] M. Handley,

[4] Assigned Numbers; J. Crowcroft, I. Wakeman. Hierarchical Rendezvous
Point proposal, work in progress.
(http://www.cs.ucl.ac.uk/staff/M.Handley/hpim.ps) Reynolds and
(ftp://cs.ucl.ac.uk/darpa/IDMR/IETF-DEC95/hpim-slides.ps).

[9] D. Estrin et al. USC/ISI, Work in progress.
(http://netweb.usc.edu/pim/).

[10] D. Estrin et al. PIM Sparse Mode Specification. Working draft,
July 1996. (draft-ietf-idmr-pim-sparse-spec-04.{ps,txt}).

[11] A. Ballardie. J. Postel; RFC 1700, October
1994.

[5] CBT - Dense Mode Interoperability: Border Router
Specification; Specification for Interconnecting a CBT Stub
Region to a DVMRP Backbone; A. Ballardie;
ftp://ds.internic.net/internet-drafts/draft-ietf-idmr-cbt-
dvmrp-**.txt. Working draft, March 1997.

[6] Scalable Multicast Key Distribution; A. Ballardie; RFC 1949, July
1996. Also available from:
ftp://cs.ucl.ac.uk/darpa/IDMR/draft-ietf-idmr-cbt-dm-interop-XX.txt

[12] S. Deering. Private communication, August 1996.

[7] A Dynamic Bootstrap Mechanism for Rendezvous-based Multicast Rout-
ing; D. Estrin et al.; Technical Report; ftp://catarina.usc.edu/pim

Author Information:

Tony Ballardie,
Research Consultant,

e-mail: ABallardie@acm.org