Internetworking Over NBMA James V. Luciani
INTERNET-DRAFT (Bay Networks)
<draft-ietf-ion-scsp-01.txt> Grenville Armitage
(Bellcore)
Joel Halpern
(Newbridge)
Expires September 1997
Server Cache Synchronization Protocol (SCSP)
Status of this Memo
This document is an Internet-Draft. Internet-Drafts are working
documents of the Internet Engineering Task Force (IETF), its areas,
and its working groups. Note that other groups may also distribute
working documents as Internet-Drafts.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as ``work in progress.''
To learn the current status of any Internet-Draft, please check the
``1id-abstracts.txt'' listing contained in the Internet-Drafts Shadow
Directories on ds.internic.net (US East Coast), nic.nordu.net
(Europe), ftp.isi.edu (US West Coast), or munnari.oz.au (Pacific
Rim).
Abstract
This document describes the Server Cache Synchronization Protocol
(SCSP) and is written in terms of SCSP's use within Non Broadcast
Multiple Access (NBMA) networks; although, a somewhat straight
forward usage is applicable to BMA networks. SCSP attempts to solve
the generalized cache synchronization/cache-replication problem for
distributed protocol entities. However, in this document, SCSP is
couched in terms of the client/server paradigm in which distributed
server entities, which are bound to a Server Group (SG) through some
means, wish to synchronize the contents (or a portion thereof) of
their caches which contain information about the state of clients
being served.
Luciani, et al. [Page 1]
INTERNET-DRAFT SCSP Expires September 1997
1. Introduction
It is perhaps an obvious goal for any protocol to not limit itself to
a single point of failure such as having a single server in a
client/server paradigm. Even when there are redundant servers, there
still remains the problem of cache synchronization; i.e., when one
server becomes aware of a change in state of cache information then
that server must propagate the knowledge of the change in state to
all servers which are actively mirroring that state information.
Further, this must be done in a timely fashion without putting undo
resource strains on the servers. Assuming that the state information
kept in the server cache is the state of clients of the server, then
in order to minimize the burden placed upon the client it is also
highly desirable that clients need not have complete knowledge of all
servers which they may use. However, any mechanism for
synchronization should not preclude a client from having access to
several (or all) servers. Of course, any solution must be reasonably
scalable, capable of using some auto-configuration service, and lend
itself to a wide range of authentication methodologies.
This document describes the Server Cache Synchronization Protocol
(SCSP). SCSP solves the generalized server synchronization/cache-
replication problem while addressing the issues described above.
SCSP synchronizes caches (or a portion of the caches) of a set of
server entities of a particular protocol which are bound to a Server
Group (SG) through some means (e.g., all NHRP servers belonging to a
Logical IP Subnet (LIS)[1]). The client/server protocol which a
particular server uses is identified by a Protocol ID (PID). SGs are
identified by an ID which, not surprisingly, is called a SGID. Note
therefore that the combination PID/SGID identifies both the
client/server protocol for which the servers of the SG are being
synchronized as well as the instance of that protocol. This implies
that multiple instances of the same protocol may be in operation at
the same time and have their servers synchronized independently of
each other. SGs may exist in any topology as long as the resultant
graph spans the set of servers that need to be synchronized. The
caches which are to be synchronized contain information on the state
of the clients within the scope of interest of the SG. An example of
types of information that must be synchronized can be seen in NHRP[2]
using IP where the information includes the REGISTERED clients' IP to
NBMA mappings in the SG LIS.
The SCSP specification is not useful as a stand alone protocol. It
must be coupled with the use of an SCSP Protocol Specific
specification which defines how a given protocol would make use of
the synchronization primitives supplied by SCSP. Such specification
will be done in separate documents; e.g., [8][9].
Luciani, et al. [Page 2]
INTERNET-DRAFT SCSP Expires September 1997
2. Overview
SCSP places no topological requirements upon the SG. Obviously,
however, the resultant graph must span the set of servers to be
synchronized. SCSP borrows its cache distribution mechanism from the
link state protocols [3,4]. However, unlike those technologies,
there is no mandatory Shortest Path First (SPF) calculation, and SCSP
imposes no additional memory requirements above and beyond that which
is required to save the cached information which would exist
regardless of the synchronization technology.
In order to give a frame of reference for the following discussion,
the terms Local Server (LS), Directly Connected Server (DCS), and
Remote Server (RS) are introduced. The LS is the server under
scrutiny; i.e., all statements are made from the perspective of the
LS when discussing the SCSP protocol. The DCS is a server which is
directly connected to the LS; e.g., there exists a VC between the LS
and DCS. Thus, every server is a DCS from the point of view of every
other server which connects to it directly, and every server is an LS
which has zero or more DCSs directly connected to it. From the
perspective of an LS, an RS is a server, separate from the LS, which
is not directly connected to the LS (i.e., an RS is always two or
more hops away from an LS whereas a DCS is always one hop away from
an LS).
SCSP contains three sub protocols: the "Hello" protocol, the "Cache
Alignment" protocol, and the "Cache State Update" protocol. The
"Hello" protocol is used to ascertain whether a DCS is operational
and whether the connection between the LS and DCS is bidirectional,
unidirectional, or non-functional. The "Cache Alignment" (CA)
protocol allows an LS to synchronize its entire cache with that of
the cache of its DCSs. The "Cache State Update" (CSU) protocol is
used to update the state of cache entries in servers for a given SG.
Sections 2.1, 2.2, and 2.3 contain a more in-depth explanation of the
Hello, CA, and CSU protocols and the messages they use.
SCSP based synchronization is performed on a per protocol instance
basis. That is, a separate instance of SCSP is run for each instance
of the given protocol running in a given box. The protocol is
identified in SCSP via a Protocol ID and the instance of the protocol
is identified by a Server Group ID (SGID). Thus the PID/SGID pair
uniquely identify an instance of SCSP. In general, this is not an
issue since it is seldom the case that many instances of a given
protocol (which is distributed and needs cache synchronization) are
running within the same physical box. However, when this is the
case, there is a mechanism called the Family ID (described briefly in
the Hello Protocol) which enables a substantial reduction in
maintenance traffic at little real cost in terms of control. The use
Luciani, et al. [Page 3]
INTERNET-DRAFT SCSP Expires September 1997
of the Family ID mechanism, when appropriate for a given protocol
which is using SCSP, will be fully defined in the given SCSP protocol
specific specification.
+---------------+
| |
+-------@| DOWN |@-------+
| | | |
| +---------------+ |
| | @ |
| | | |
| | | |
| | | |
| @ | |
| +---------------+ |
| | | |
| | WAITING | |
| +--| |--+ |
| | +---------------+ | |
| | @ @ | |
| | | | | |
| @ | | @ |
+---------------+ +---------------+
| BIDIRECTIONAL |----@| UNIDIRECTIONAL|
| | | |
| CONNECTION |@----| CONNECTION |
+---------------+ +---------------+
Figure 1: Hello Finite State Machine (HFSM)
2.1 Hello Protocol
"Hello" messages are used to ascertain whether a DCS is operational
and whether the connections between the LS and DCS are bidirectional,
unidirectional, or non-functional. In order to do this, every LS MUST
periodically send a Hello message to its DCSs.
An LS must be configured with a list of NBMA addresses which
represent the addresses of peer servers in a SG to which the LS
wishes to have a direct connection for the purpose of running SCSP;
that is, these addresses are the addresses of would-be DCSs. The
mechanism for the configuration of an LS with these NBMA address is
beyond the scope of this document; although one possible mechanism
would be an autoconfiguration server.
Luciani, et al. [Page 4]
INTERNET-DRAFT SCSP Expires September 1997
An LS has a Hello Finite State Machine (HFSM) associated with each of
its DCSs (see Figure 1) for a given SG, and the HFSM monitors the
state of the connectivity between the servers.
The HFSM starts in the "Down" State and transitions to the "Waiting"
State after NBMA level connectivity has been established. Once in
the Waiting State, the LS starts sending Hello messages to the DCS.
The Hello message includes: a Sender ID which is set to the LS's ID
(LSID), zero or more Receiver IDs which identify the DCSs from which
the LS has heard a Hello message, and a HelloInterval and DeadFactor
which will be described below. At this point, the DCS may or may
not already be sending its own Hello messages to the LS.
When the LS receives a Hello message from one of its DCSs, the LS
checks to see if its LSID is in one of the Receiver ID fields of that
message which it just received, and the LS saves the Sender ID from
that Hello message. If the LSID is in one of the Receiver ID fields
then the LS transitions the HFSM to the Bidirectional Connection
state otherwise it transitions the HFSM into the Unidirectional
Connection state. The Sender ID which was saved is the DCS's ID
(DCSID). The next time that the LS sends its own Hello message to
the DCS, the LS will check the saved DCSID against a list of Receiver
IDs which the LS uses when sending the LS's own Hello messages. If
the DCSID is not found in the list of Receiver IDs then it is added
to that list before the LS sends its Hello message.
Hello messages also contain a HelloInterval and a DeadFactor. The
Hello interval advertises the time (in seconds) between sending of
consecutive Hello messages by the server which is sending the
"current" Hello message. That is, if the time between reception of
Hello messages from a DCS exceeds the HelloInterval advertised by
that DCS then the next Hello message is to be considered late by the
LS. If the LS does not receive a Hello message, which contains the
LS's LSID in one of the Receiver ID fields, within the interval
HelloInterval*DeadFactor seconds (where DeadFactor was advertised by
the DCS in a previous Hello message) then the LS MUST consider the
DCS to be stalled. At which point one of two things will happen: 1)
if any Hello messages have been received during the last
HelloInterval*DeadFactor seconds then the LS should transition the
HFSM for that DCS to the Unidirectional Connection State; otherwise,
the LS should transition the HFSM for that DCS to the Waiting State
and remove the DCSID from the Receiver ID list.
Note that the Hello Protocol is on a per PID/SGID basis. Thus, for
example, if there are two servers (one in SG A and the other in SG B)
associated with an NBMA address X and another two servers (also one
in SG A and the other in SG B) associated with NBMA address Y and
there is a suitable point-to-point VC between the NBMA addresses then
Luciani, et al. [Page 5]
INTERNET-DRAFT SCSP Expires September 1997
there are two HFSMs running on each side of the VC (one per
PID/SGID).
Hello messages contain a list of Receiver IDs instead of a single
Receiver ID in order to make use of point to multipoint connections.
While there is an HFSM per DCS, an LS MUST send only a single Hello
message to its DCSs attached as leaves of a point to multipoint
connection. The LS does this by including DCSIDs in the list of
Receiver IDs when the LS's sends its next Hello message. Only the
DCSIDs from non-stalled DCSs from which the LS has heard a Hello
message are included.
Any abnormal event, such as receiving a malformed SCSP message,
causes the HFSM to transition to the Waiting State; however, a loss
of NBMA connectivity causes the HFSM to transition to the Down State.
Until, the HFSM is in the Bidirectional Connection State any properly
formed SCSP messages other than Hello messages must be ignored (this
is for the case where, for example, there is a point to multipoint
connection involved).
Luciani, et al. [Page 6]
INTERNET-DRAFT SCSP Expires September 1997
+------------+
| |
+---@| DOWN |
| | |
| +------------+
| |
| |
| @
| +------------+
| |Master/Slave|
|----| |@---+
| |Negotiation | |
| +------------+ |
| | |
| | |
| @ |
| +------------+ |
| | Cache | |
|----| |----|
| | Summarize | |
| +------------+ |
| | |
| | |
| @ |
| +------------+ |
| | Update | |
|----| |----|
| | Cache | |
| +------------+ |
| | |
| | |
| @ |
| +------------+ |
| | | |
+----| Aligned |----+
| |
+------------+
Figure 2: Cache Alignment Finite State Machine
2.2 Cache Alignment Protocol
"Cache Alignment" (CA) messages are used by an LS to synchronize its
cache with that of the cache of each of its DCSs. That is, CA
messages allow a booting LS to synchronize with each of its DCSs. A
CA message contains a CA header followed by zero or more Cache State
Luciani, et al. [Page 7]
INTERNET-DRAFT SCSP Expires September 1997
Advertisement Summary records (CSAS records).
An LS has a Cache Alignment Finite State Machine (CAFSM) associated
(see Figure 2) with each of its DCSs on a per PID/SGID basis, and the
CAFSM monitors the state of the cache alignment between the servers.
The CAFSM starts in the Down State. The CAFSM is associated with an
HFSM, and when that HFSM reaches the Bidirectional State, the CAFSM
transitions to the Master/Slave Negotiation State. The Master/Slave
Negotiation State causes either the LS or DCS to take on the role of
master over the cache alignment process.
When the LS's CAFSM reaches the Master/Slave Negotiation State, the
LS will send a CA message to the DCS associated with the CAFSM. The
format of CA messages are described in Section B.2.1. The first CA
message which the LS sends includes no CSAS records and a CA header
which contains the LSID in the Sender ID field, the DCSID in the
Receiver ID field, a CA sequence number, and three bits. These three
bits are the M (Master/Slave) bit, the I (Initialization of master)
bit, and the O (More) bit. In the first CA message sent by the LS to
a particular DCS, the M, O, and I bits are set to one. If the LS
does not receive a CA message from the DCS in CAReXmtInterval seconds
then it resends the CA message it just sent. The LS continues to do
this until the CAFSM transitions to the Cache Summarize State or
until the HFSM transitions out of the Bidirectional State. Any time
the HFSM transitions out of the Bidirectional State, the CAFSM
transitions to the Down State.
2.2.1 Master Slave Negotiation State
When the LS receives a CA message from the DCS while in the
Master/Slave Negotiation State, the role the LS plays in the exchange
depends on packet processing as follows:
1) If the CA from the DCS has the M, I, and O bits set to one and there are
no CSAS records in the CA message and the Sender ID as specified in the
DCS's CA message is larger than the LSID then
a) The timer counting down the CAReXmtInterval is stopped.
b) The CAFSM corresponding to that DCS transitions to the Cache Summarize
State and the LS takes on the role of slave.
c) The LS adopts the CA sequence number it received in the CA message
as its
own CA sequence number.
d) The LS sends a CA message to the DCS which is formated as follows:
the M and I bits are set to zero, the Sender ID field is set to the
LSID, the Receiver ID field is set to the DCSID, and the CA sequence
number is set to the CA sequence number that appeared in the DCS's
CA message. If there are CSAS records to be sent (i.e., if the LS's
cache is not empty) then the O bit is set to one and the initial set
of CSAS records are included in the CA message.
Luciani, et al. [Page 8]
INTERNET-DRAFT SCSP Expires September 1997
2) If the CA message from the DCS has the M and I bits off and the
Sender ID
as specified in the DCS's CA message is smaller than the LSID then
a) The timer counting down the CAReXmtInterval is stopped.
b) The CAFSM corresponding to that DCS transitions to the Cache Summarize
State and the LS takes on the role of master.
c) The LS must process the received CA message.
An explanation of CA message processing is given below.
d) The LS sends a CA message to the DCS which is formated as follows:
the M bit is set to one, I bit is set to zero, the Sender ID
field is set to the LSID, the Receiver ID field is set to the DCSID,
and the LS's current CA sequence number is incremented by one and
placed
in the CA message. If there are any CSAS records to be sent from the
LS to the DCS (i.e., if the LS's cache is not empty) then the O bit is
set to one and the initial set of CSAS records are included in the
CA message that the LS is sending to the DCS.
3) Otherwise, the packet must be ignored.
2.2.2 The Cache Summarize State
At any given time, the master or slave have at most one outstanding
CA message. Once the LS's CAFSM has transitioned to the Cache
Summarize State the sequence of exchanges of CA messages occurs as
follows.
1) If the LS receives a CA message with the M bit set incorrectly
(e.g., the M bit is set in the CA of the DCS and the LS is master)
or if the I bit is set then the CAFSM transitions back to the
Master/Slave Negotiation State.
2) If the LS is master and the LS receives a CA message with a CA sequence
number which is one less than the LS's current CA sequence number then
the message is a duplicate and the message MUST be discarded.
3) If the LS is master and the LS receives a CA message with a CA sequence
number which is equal to the LS's current CA sequence number then the
CA message MUST be processed. An explanation of "CA message processing"
is given below. As a result of having received the CA message from
the DCS the following will occur:
a) The timer counting down the CAReXmtInterval is stopped.
b) The LS must process any CSAS records in the received CA message.
c) Increment the LS's CA sequence number by one.
d) The cache exchange continues as follows:
1) If the LS has no more CSAS records to send and the received CA
message has the O bit off then the CAFSM transitions to the Update
Cache State.
2) If the LS has no more CSAS records to send and the received CA
message has the O bit on then the LS sends back a CA message
Luciani, et al. [Page 9]
INTERNET-DRAFT SCSP Expires September 1997
(with new CA sequence number) which contains no CSAS records and
with the O bit off. Reset the timer counting down the
CAReXmtInterval.
3) If the LS has more CSAS records to send then the LS sends the next
CA message with the LS's next set of CSAS records. If LS is sending
its last set of CSAS records then the O bit is set off otherwise the
O bit is set on. Reset the timer counting down the CAReXmtInterval.
4) If the LS is slave and the LS receives a CA message with a CA sequence
number which is equal to the LS's current CA sequence number then the
CA message is a duplicate and the LS MUST resend the CA message
which it had just sent to the DCS.
5) If the LS is slave and the LS receives a CA message with a CA sequence
number which is one more than the LS's current CA sequence number then
the message is valid and MUST be processed. An explanation of "CA
message
processing" is given below. As a result of having received the CA
message from the DCS the following will occur:
a) The LS must process any CSAS records in the received CA message.
b) Set the LS's CA sequence number to the CA sequence number in the CA
message.
c) The cache exchange continues as follows:
1) If the LS had just sent a CA message with the O bit off and the
received CA message has the O bit off then the CAFSM transitions to
the Update Cache State and the LS sends a CA message with no CSAS
records and with the O bit off.
2) If the LS still has CSAS records to send then the LS MUST send
a CA message with CSAS records in it.
a) If the message being sent from the LS to the DCS does not contain
the last CSAS records that the LS needs to send then the CA
message is sent with the O bit on.
b) If the message being sent from the LS to the DCS does contain
the last CSAS records that the LS needs to send and
the CA message just received from the DCS had the O bit off then
the CA message is sent with the O bit off, and the LS transitions
the CAFSM to the Update Cache State.
c) If the message being sent from the LS to the DCS does contain
the last CSAS records that the LS needs to send and
the CA message just received from the DCS had the O bit on then
the CA message is sent with the O bit off and the alignment
process continues.
6) If the LS is slave and the LS receives a CA message with a CA sequence
number that is neither equal to nor one more than the current LS's
CA sequence number then an error has occurred and the CAFSM transitions
to the Master/Slave Negotiation State.
Luciani, et al. [Page 10]
INTERNET-DRAFT SCSP Expires September 1997
Note that if the LS was slave during the CA process then the LS upon
transitioning the CAFSM to the Update Cache state MUST keep a copy of
the last CA message it sent and set a timer equal to CAReXmtInterval.
If either the timer expires or the LS receives a CSU Solicit (CSUS)
message (CSUS messages are described in Section 2.2.3) from the DCS
then the LS releases the copy of the CA message. The reason for this
is that if the DCS (which is master) loses the last CA message sent
by the LS then the DCS will resend its previous CA message with the
last CA Sequence number used. If that were to occur the LS would
need to resend its last sent CA message as well.
2.2.2.1 "CA message processing":
The LS makes a list of those cache entries which are more "up to
date" in the DCS than the LS's own cache. This list is called the
CSA Request List (CRL). See Section 2.4 for a description of what it
means for a CSA (Client State Advertisement) record or CSAS record to
be more "up to date" than an LS's cache entry.
2.2.3 The Update Cache State
If the CRL of the associated CAFSM of the LS is empty upon transition
into the Update Cache State then the CAFSM immediately transitions
into the Aligned State.
If the CRL is not empty upon transition into the Update Cache State
then the LS solicits the DCS to send the CSA records corresponding to
the summaries (i.e., CSAS records) which the LS holds in its CRL. The
solicited CSA records will contain the entirety of the cached
information held in the DCS's cache for the given cache entry. The
LS solicits the relevant CSA records by forming CSU Solicit (CSUS)
messages from the CRL. See Section B.2.4 for the description of the
CSUS message format. The LS then sends the CSUS messages to the DCS.
The DCS responds to the CSUS message by sending to the LS one or more
CSU Request messages containing the entirety of newer cached
information identified in the CSUS message. Upon receiving the CSU
Request the LS will send one or more CSU Replies as described in
Section 2.3. Note that the LS may have at most one CSUS message
outstanding at any given time.
Just before the first CSUS message is sent from an LS to the DCS
associated with the CAFSM, a timer is set to CSUSReXmtInterval
seconds. If all the CSA records corresponding to the CSAS records in
the CSUS message have not been received by the time that the timer
expires then a new CSUS message will be created which contains all
the CSAS records for which no appropriate CSA record has been
received plus additional CSAS records not covered in the previous
CSUS message. The new CSUS message is then sent to the DCS. If, at
Luciani, et al. [Page 11]
INTERNET-DRAFT SCSP Expires September 1997
some point before the timer expires, all CSA record updates have been
received for all the CSAS records included in the previously sent
CSUS message then the timer is stopped. Once the timer is stopped,
if there are additional CSAS records that were not covered in the
previous CSUS message but were in the CRL then the timer is reset and
a new CSUS message is created which contains only those CSAS records
from the CRL which have not yet been sent to the DCS. This process
continues until all the CSA records corresponding CSAS records that
were in the CRL have been received by the LS. When the LS has a
completely updated cache then the LS transitions CAFSM associated
with the DCS to the Aligned State.
If an LS receives a CSUS message or a CA message with a Receiver ID
which is not the LS's LSID then the message must be discarded and
ignored. This is necessary since an LS may be a leaf of a point to
multipoint connection with other servers in the SG.
2.2.4 The Aligned State
While in the Aligned state, an LS will perform the Cache State Update
Protocol as described in Section 2.3.
Note that an LS may receive a CSUS message while in the Aligned State
and, the LS MUST respond to the CSUS message with the appropriate CSU
Request message in a similar fashion to the method previously
described in Section 2.2.3.
2.3 Cache State Update Protocol
"Cache State Update" (CSU) messages are used to dynamically update
the state of cache entries in servers on a given PID/SGID basis. CSU
messages contain zero or more "Cache State Advertisement" (CSA)
records each of which contains its own snapshot of the state of a
particular cache entry. An LS may send/receive a CSU to/from a DCS
only when the corresponding CAFSM is in either the Aligned State or
the Update Cache State.
There are two types of CSU messages: CSU Requests and CSU Replies.
See Sections B.2.2 and B.2.3 respectively for message formats. A CSU
Request message is sent from an LS to one or more DCSs for one of two
reasons: either the LS has received a CSUS message and MUST respond
only to the DCS which originated the CSUS message, or the LS has
become aware of a change of state of a cache entry. An LS becomes
aware of a change of state of a cache entry either through receiving
a CSU Request from one of its DCSs or as a result of a change of
state being observed in a cached entry originated by the LS. In the
former case, the LS will send a CSU Request to each of its DCSs
Luciani, et al. [Page 12]
INTERNET-DRAFT SCSP Expires September 1997
except the DCS from which the LS became aware of the change in state.
In the latter case, the LS will send a CSU Request to each of its
DCSs. The change in state of a particular cache entry is noted in a
CSA record which is then appended to the end of the CSU Request
message mandatory part. In this way, state changes are propagated
throughout the SG.
Examples of such changes in state are as follows:
1) an server receives a request from a client to add an entry to its
cache,
2) an server receives a request from a client to remove an entry
from its
cache
3) a cache entry has timed out in the server's cache, has been
refreshed
in the server's cache, or has been administratively modified
When an LS receives a CSU Request from one of its DCSs, the LS
acknowledges one or more CSA Records which were contained in the CSU
Request by sending a CSU Reply. The CSU Reply contains one or more
CSAS records which correspond to those CSA records which are being
acknowledged. Thus, for example, if a CSA record is dropped (or
delayed in processing) by the LS because there are insufficient
resources to process it then a corresponding CSAS record is not
included in the CSU Reply to the DCS.
Note that an LS may send multiple CSU Request messages before
receiving a CSU Reply acknowledging any of the CSA Records contained
in the CSU Requests. Note also that a CSU Reply may contain
acknowledgments for CSA Records from multiple CSU Requests. Thus,
the terms "request" and "reply" may be a bit confusing.
Note that a CSA Record contains a CSAS Record followed by
client/server protocol specific information contained in a cache
entry (see Section B.2.0.2 for CSAS record format information and
Section B.2.2.1 for CSA record format information). When a CSA
record is considered by the LS to represent cached information which
is more "up to date" (see Section 2.4) than the cached information
contained within the cache of the LS then two things happen: 1) the
LS's cache is updated with the more up to date information, and 2)
the LS sends a CSU Request containing the CSA Record to each of its
DCSs except the one from which the CSA Record arrived. In this way,
state changes are propagated within the PID/SGID. Of course, at some
point, the LS will also acknowledge the reception of the CSA Record
by sending the appropriate DCS a CSU Reply message containing the
corresponding CSAS Record.
When an LS sends a new CSU Request, the LS keeps track of the
outstanding CSA records in that CSU Request and to which DCSs the LS
sent the CSU Request. For each DCS to which the CSU Request was
Luciani, et al. [Page 13]
INTERNET-DRAFT SCSP Expires September 1997
sent, a timer set to CSUReXmtInterval seconds is started just prior
to sending the CSU Request. This timer is associated with the CSA
Records contained in that CSU Request such that if that timer expires
prior to having all CSA Records acknowledged from that DCS then (and
only then) a CSU Request is re-sent by the LS to that DCS. However,
the re-sent CSU Request only contains those CSA Records which have
not yet been acknowledged. If all CSA Records associated with a
timer becomes acknowledged then the timer is stopped. Note that the
re-sent CSA Records follow the same time-out and retransmit rules as
if they were new. Retransmission will occur a configured number of
times for a given CSA Record and if acknowledgment fails to occur
then an "abnormal event" has occurred at which point the then the
HFSM associated with the DCS is transitioned to the Waiting State.
A CSA Record instance is said to be on a "DCS retransmit queue" when
it is associated with the previously mentioned timer. Only the most
up-to-date CSA Record is permitted to be queued to a given DCS
retransmit queue. Thus, if a less up-to-date CSA Record is queued to
the DCS retransmit queue when a newer CSA Record instance is about to
be queued to that DCS retransmit queue then the older CSA Record
instance is dequeued and disassociated with its timer immediately
prior to enqueuing the newer instance of the CSA Record.
When an LS receives a CSU Reply from one of its DCSs then the LS
checks each CSAS record in the CSU Reply against the CSAS Record
portion of the CSA Records which are queued to the DCS retransmit
queue.
1) If there exists an exact match between the CSAS record portion
of the CSA record and a CSAS Record in the CSU Reply then
that CSA Record is considered to be acknowledged and is thus
dequeued from the DCS retransmit queue and is
disassociated with its timer.
2) If there exists a match between the CSAS record portion
of the CSA record and a CSAS Record in the CSU Reply except
for the CSA Sequence number then
a) If the CSA Record queued to the DCS retransmit queue has a
CSA Sequence Number which is greater than the
CSA Sequence Number in the CSAS Record of the the CSU Reply then
the CSAS Record in the CSU Reply is ignored.
b) If the CSA Record queued to the DCS retransmit queue has a
CSA Sequence Number which is less than the
CSA Sequence Number in the CSAS Record of the the CSU Reply then
CSA Record which is queued to the DCS retransmit queue is
dequeued and the CSA Record is disassociated with its timer.
Further, a CSUS Message is sent to that DCS which sent the
more up-to-date CSAS Record. All normal CSUS processing
Luciani, et al. [Page 14]
INTERNET-DRAFT SCSP Expires September 1997
occurs as if the CSUS were sent as part of the CA protocol.
When an LS receives a CSU Request message which contains a CSA Record
which contains a CSA Sequence Number which is smaller than the CSA
Sequence number of the cached CSA then the LS MUST acknowledge the
CSA record in the CSU Request but it MUST do so by sending a CSU
Reply message containing the CSAS Record portion of the CSA Record
stored in the cache and not the CSAS Record portion of the CSA Record
contained in the CSU Request.
An LS responds to CSUS messages from its DCSs by sending CSU Request
messages containing the appropriate CSA records to the DCS. If an LS
receives a CSUS message containing a CSAS record for an entry which
is no longer in its database (e.g., the entry timed out and was
discarded after the Cache Alignment exchange completed but before the
entry was requested through a CSUS message), then the LS will respond
by copying the CSAS Record from the CSUS message into a CSU Request
message and the LS will set the N bit signifying that this record is
a NULL record since the cache entry no longer exists in the LS's
cache. Note that in this case, the "CSA Record" included in the CSU
Request to signify the NULL cache entry is literally only a CSAS
Record since no client/server protocol specific information exists
for the cache entry.
If an LS receives a CSA Record in a CSU Request from a DCS for which
the LS has an identical CSA record posted to the corresponding DCS's
DCS retransmit queue then the CSA Record on the DCS retransmit queue
is considered to be implicitly acknowledged. Thus, the CSA Record is
dequeued from the DCS retransmit queue and is disassociated with its
timer. The CSA Record sent by the DCS MUST still be acknowledged by
the LS in a CSU Reply, however. This is useful in the case of point
to multipoint connections where the rule that "when an LS receives a
CSA record from a DCS, that LS floods the CSA Record to every DCS
except the DCS from which it was received" might be broken.
If an LS receives a CSU with a Receiver ID which is not equal to the
LSID and is not set to all 0xFFs then the CSU must be discarded and
ignored. This is necessary since the LS may be a leaf of a point to
multipoint connection with other servers in the LS's SG.
An LS MAY send a CSU Request to the all 0xFFs Receiver ID when the LS
is a root of a point to multipoint connection with a set of its DCSs.
If an LS receives a CSU Request with the all 0xFFs Receiver ID then
it MUST use the Sender ID in the CSU Request as the Receiver ID of
the CSU Reply (i.e., it MUST unicast its response to the sender of
the request) when responding. If the LS wishes to send a CSU Request
to the all 0xFFs Receiver ID then it MUST create a time-out and
retransmit timer for each of the DCSs which are leaves of the point
Luciani, et al. [Page 15]
INTERNET-DRAFT SCSP Expires September 1997
to multipoint connection prior to sending the CSU Request. If in
this case, the time-out and retransmit timer expires for a given DCS
prior to acknowledgment of a given CSA Record then the LS MUST use
the specific DCSID as the Receiver ID rather than the all 0xFFs
Receiver ID. Similarly, if it is necessary to re-send a CSA Record
then the LS MUST specify the specific DCSID as the Receiver ID rather
than the all 0xFFs Receiver ID.
Note that if a set of servers are in a full mesh of point to
multipoint connections, and one server of that mesh sends a CSU
Request into that full mesh, and the sending server sends the CSA
Records in the CSU Request to the all 0xFFs Receiver ID then it would
not be necessary for every other server in the mesh to source their
own CSU Request containing those CSA Records into the mesh in order
to properly flood the CSA Records. This is because every server in
the mesh would have heard the CSU Request and would have processed
the included CSA Records as appropriate. Thus, a server in a full
mesh could consider the mesh to be a single logical port and so the
rule that "when an LS receives a CSA record from a DCS, that LS
floods the CSA Record to every DCS except the DCS from which it was
received" is not broken. A receiving server in the full mesh would
still need to acknowledge the CSA records with CSU Reply messages
which contain the LSID of the replying server as the Sender ID and
the ID of the server which sent the CSU Request as the Receiver ID
field. In the time out and retransmit case, the Receiver ID of the
CSU Request would be set to the specific DCSID which did not
acknowledge the CSA Record (as opposed to the all 0xFFs Receiver ID).
Since a full mesh emulates a broadcast media for the servers attached
to the full mesh, use of SCSP on a broadcast medium might use this
technique as well. Further discussion of this use of a full mesh or
use of a broadcast media is left to the client/server protocol
specific documents.
2.4 The meaning of "More Up To Date"/"Newness"
During the cache alignment process and during normal CSU processing,
a CSAS Record is compared against the contents of an LS's cache entry
to decide whether the information contained in the record is more "up
to date" than the corresponding cache entry of the LS.
There are three pieces of information which are used in determining
whether a record contains information which is more "up to date" than
the information contained in the cache entry of an LS which is
processing the record: 1) the Cache Key, 2) the Originator which is
described by an Originator ID (OID), and 3) the CSA Sequence number.
See Section B.2.0.2 for more information on these fields.
Luciani, et al. [Page 16]
INTERNET-DRAFT SCSP Expires September 1997
Given these three pieces of information, a CSAS record (be it part of
a CSA Record or be it stand-alone) is considered to be more "up to
date" than the information contained in the cache of an LS if all of
the following are true:
1) The Cache Key in the CSAS Record matches the stored Cache Key
in the LS's cache entry,
2) The OID in the CSAS Record matches the stored OID
in the LS's cache entry,
3) The CSA Sequence Number in the CSAS Record is greater than
CSA Sequence Number in the LS's cache entry.
Discussion and conclusions
While the above text is couched in terms of synchronizing the
knowledge of the state of a client within the cache of servers
contained in a SG, this solution generalizes easily to any number of
database synchronization problems (e.g., LECS synchronization).
SCSP defines a generic flooding protocol. There are a number of
related issues relative to cache maintenance and topology maintenance
which are more appropriately defined in the client/server protocol
specific documents; for example, it might be desirable to define a
generic cache entry time-out mechanism for a given protocol or to
advertise adjacency information between servers so that one could
obtain a topo-map of the servers in a SG. When mechanisms like these
are desirable, they will be defined in the client/server protocol
specific documents.
Appendix A: Terminology and Definitions
CA Message - Cache Alignment Message
These messages allow an LS to synchronize its entire cache with
that of the cache of one of its DCSs.
CAFSM - Cache Alignment Finite State Machine
The CAFSM monitors the state of the cache alignment between an LS
and a particular DCS. There exists one CAFSM per DCS as seen from
an LS.
CSA Record - Cache State Advertisement Record
A CSA is a record within a CSU message which identifies an update
to the status of a "particular" cache entry.
CSAS Record - Cache State Advertisement Summary Record
A CSAS contains a summary of the information in a CSA. A server
will send CSAS records describing its cache entries to another
Luciani, et al. [Page 17]
INTERNET-DRAFT SCSP Expires September 1997
server during the cache alignment process. CSAS records are also
included in a CSUS messages when an LS wants to request the entire
CSA from the DCS. The LS is requesting the CSA from the DCS
because the LS believes that the DCS has a more recent view of the
state of the cache entry in question.
CSU Message - Cache State Update message
This is a message sent from an LS to its DCSs when the LS becomes
aware of a change in state of a cache entry.
CSUS Message - Cache State Update Solicit Message
This message is sent by an LS to its DCS after the LS and DCS have
exchanged CA messages. The CSUS message contains one or more CSAS
records which represent solicitations for entire CSA records (as
opposed to just the summary information held in the CSAS).
DCS - Directly Connected Server
The DCS is a server which is directly connected to the LS; e.g.,
there exists a VC between the LS and DCS. This term, along with the
terms LS and RS, is used to give a frame of reference when talking
about servers and their synchronization. Unless explicitly stated
to the contrary, there is no implied difference in functionality
between a DCS, LS, and RS.
HFSM - Hello Finite State Machine
An LS has a HFSM associated with each of its DCSs. The HFSM
monitors the state of the connectivity between the LS and a
particular DCS.
LS - Local Server
The LS is the server under scrutiny; i.e., all statements are made
from the perspective of the LS. This term, along with the terms
DCS and RS, is used to give a frame of reference when talking about
servers and their synchronization. Unless explicitly stated to the
contrary, there is no implied difference in functionality between a
DCS, LS, and RS.
LSID - Local Server ID
The LSID is a unique token that identifies an LS. This value might
be taken from the protocol address of the LS.
PID - Protocol ID
This field contains an identifier which identifies the
client/server protocol which is making use of SCSP for the given
message. The assignment of Protocol IDs for this field is given
over to IANA.
RS - Remote Server (RS)
Luciani, et al. [Page 18]
INTERNET-DRAFT SCSP Expires September 1997
From the perspective of an LS, an RS is a server, separate from the
LS, which is not directly connected to the LS (i.e., an RS is
always two or more hops away from an LS whereas a DCS is always one
hop away from an LS). Unless otherwise stated an RS refers to a
server in the SG. This term, along with the terms LS and DCS, is
used to give a frame of reference when talking about servers and
their synchronization. Unless explicitly stated to the contrary,
there is no implied difference in functionality between a DCS, LS,
and RS.
SG - Server Group
The SCSP synchronizes caches (or a portion of the caches) of a set
of server entities which are bound to a SG through some means
(e.g., all servers belonging to a Logical IP Subnet (LIS)[1]).
Thus an SG is just a grouping of servers around some commonality.
SGID - Server Group ID
This ID is a 16 bit identification field that uniquely identifies
the instance client/server protocol for which the servers of the SG
are being synchronized. This implies that multiple instances of
the same protocol may be in operation at the same time and have
their servers synchronized independently of each other.
Appendix B: SCSP Message Formats
This section of the appendix includes the message formats for SCSP.
SCSP protocols are LLC/SNAP encapsulated with an LLC=0xAA-AA-03 and
OUI=0x00-00-5e and PID=0x00-05.
SCSP has 3 parts to every packet: the fixed part, the mandatory part,
and the extensions part. The fixed part of the message exists in
every packet and is shown below. The mandatory part is specific to
the particular message type (i.e., CA, CSU Request/Reply, Hello,
CSUS) and, it includes (among other packet elements) a Mandatory
Common Part and zero or more records each of which contains
information pertinent to the state of a particular cache entry
(except in the case of a Hello message) whose information is being
synchronized within a SG. The extensions part contains the set of
extensions for the SCSP message.
In the following message formats, "unused" fields are set to zero on
when transmitting a message and these fields are ignored on receipt
of a message.
B.1 Fixed Part
Luciani, et al. [Page 19]
INTERNET-DRAFT SCSP Expires September 1997
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Version | Type Code | Packet Size |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Checksum | Start Of Extensions |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Version
This is the version of the SCSP protocol being used. The current
version is 1.
Type Code
This is the code for the message type (e.g., Hello (5), CSU
Request(2), CSU Reply(3), CSUS (4), CA (1)).
Packet Size
The total length of the SCSP packet, in octets (excluding link
layer and/or other protocol encapsulation).
Checksum
The standard IP checksum over the entire SCSP packet (starting with
the fixed header).
Start Of Extensions
This field is coded as zero when no extensions are present in the
message. If extensions are present then this field will be coded
with the offset from the top of the fixed header to the beginning
of the first extension.
B.2.0 Mandatory Part
The mandatory part of the SCSP packet contains the operation specific
information for a given message type (e.g., SCSP Cache State Update
Request/Reply, etc.), and it includes (among other packet elements) a
Mandatory Common Part (described in Section B.2.0.1) and zero or more
records each of which contains information pertinent to the state of
a particular cache entry (except in the case of a Hello message)
whose information is being synchronized within a SG. These records
may, depending on the message type, be either Cache State
Advertisement Summary (CSAS) Records (described in Section B.2.0.2)
or Cache State Advertisement (CSA) Records (described in Section
B.2.2.1). CSA Records contain a summary of a cache entry's
information (i.e., a CSAS Record) plus some additional client/server
protocol specific information. The mandatory common part format and
CSAS Record format is shown immediately below, prior to showing their
use in SCSP messages, in order to prevent replication within the
Luciani, et al. [Page 20]
INTERNET-DRAFT SCSP Expires September 1997
message descriptions.
B.2.0.1 Mandatory Common Part
Sections B.2.1 through B.2.5 have a substantial overlap in format.
This overlapping format is called the mandatory common part and its
format is shown below:
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Protocol ID | Server Group ID |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| unused | Flags |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Sender ID Len | Recvr ID Len | Number of Records |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Sender ID (variable length) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Receiver ID (variable length) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Protocol ID
This field contains an identifier which identifies the
client/server protocol which is making use of SCSP for the given
message. The assignment of Protocol IDs for this field is given
over to IANA. IANA will accept any and all requests for value
assignment as long as the client/server protocol specific document
exists. Protocols with current documents have the the following
defined values:
1 - ATMARP
2 - NHRP
3 - MARS
4 - DHCP
5 - LNNI
Server Group ID
This ID is uniquely identifies the instance of a given
client/server protocol for which servers are being synchronized.
Flags
The Flags field is message specific, and its use will be described
in the specific message format sections below.
Sender ID Len
This field holds the length in octets of the Sender ID.
Recvr ID Len
This field holds the length in octets of the Receiver ID.
Luciani, et al. [Page 21]
INTERNET-DRAFT SCSP Expires September 1997
Number of Records
This field contains the number of additional records associated
with the given message. The exact format of these records is
specific to the message and will be described for each message type
in the sections below.
Sender ID
This is an identifier assigned to the server which is sending the
given message. One possible assignment might be the protocol
address of the sending server.
Receiver ID
This is an identifier assigned to the server which is to receive
the given message. One possible assignment might be the protocol
address of the server which is to receive the given message.
B.2.0.2 Cache State Advertisement Summary Record (CSAS record)
CSAS records contain a summary of information contained in a cache
entry of a given client/server database which is being synchronized
through the use of SCSP. The summary includes enough information for
SCSP to look into the client/server database for the appropriate
database cache entry and then compare the "newness" of the summary
against the "newness" of the cached entry.
Note that CSAS records do not contain a Server Group ID (SGID) nor do
they contain a Protocol ID. These IDs are necessary to identify
which protocol and which instance of that protocol for which the
summary is applicable. These IDs are present in the mandatory common
part of each message.
Note also that the values of the Hop Count and Record Length fields
of a CSAS Record are dependent on whether the CSAS record exists as a
"stand-alone" record or whether the CSAS record is "embedded" in CSA
Record. This is further described below.
Luciani, et al. [Page 22]
INTERNET-DRAFT SCSP Expires September 1997
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Hop Count | Record Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cache Key Len | Orig ID Len |N| unused |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CSA Sequence Number |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cache Key ... |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Originator ID ... |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Hop Count
This field represents the number of hops that the record may take
before being dropped. Thus, at each server that the record
traverses, the Hop Count is decremented. This field is set to 1
when the CSAS record is a "stand-alone" record (i.e., it is not
embedded within a CSA record) since summaries do not go beyond one
hop during the cache alignment process. If a CSAS record is
"embedded" within a CSA record then the Hop Count is set to an
administratively defined value which is almost certainly
proportional to the "diameter" of the SG being synchronized.
Record Length
If the CSAS record is a "stand-alone" record then this value is
12+"Cache Key Leng"+"Orig ID Len" in bytes; otherwise, this value
is set to 12+"Cache Key Leng"+"Orig ID Len"+ sizeof("Client/Server
Protocol Specific Part for cache entry"). The size of the
Client/Server Protocol Specific Part may be obtained from the
client/server protocol specific document for the given Protocol ID.
Cache Key Len
Length of the Cache Key field in bytes.
Orig ID Len.
Length of the Originator ID field in bytes.
N
The "N" bit signifies that this CSAS Record is actually a Null
record. This bit is only used in a CSAS Record contained in a CSU
Request/Reply which is sent in response to a CSUS message. It is
possible that an LS may receive a solicitation for a CSA record
when the cache entry represented by the solicited CSA Record no
longer exists in the LS's cache (see Section 2.3 for details). In
this case, the LS copies the CSAS Record directly from the CSUS
message into the CSU Request, and the LS sets the N bit signifying
Luciani, et al. [Page 23]
INTERNET-DRAFT SCSP Expires September 1997
that the cache entry does not exist any longer. The DCS which
solicited the CSA record which no longer exists will still respond
with a CSU Reply. This bit is usually set to zero.
CSA Sequence Number
This field contains a sequence number that identifies the "newness"
of a CSA record instance being summarized. A "larger" sequence
number means a more recent advertisement. Thus, if the state of
part (or all) of a cache entry needs to be updated then the CSA
record advertising the new state MUST contain a CSA Sequence Number
which is larger than the one corresponding to the previous
advertisement. This number is assigned by the originator of the
CSA record. The CSA Sequence Number may be assigned by the
originating server or by the client which caused its server to
advertise its existence.
Cache Key
This is a database lookup key that uniquely identifies a piece of
data which the originator of a CSA Record wishes to synchronize
with its peers for a given "Protocol ID/Server Group ID" pair.
This key will generally be a small opaque byte string which SCSP
will associate with a given piece of data in a cache. Thus, for
example, an originator might assign a particular 4 byte string to
the binding of an IP address with that of an ATM address.
Generally speaking, the originating server of a CSA record is
responsible for generating a Cache Key for every element of data
that the the given server originates and which the server wishes to
synchronize with its peers in the SG.
Originator ID
This field contains an ID administratively assigned to the server
which is the originator of CSA Records.
B.2.1 Cache Alignment (CA)
The Cache Alignment (CA) message allows an LS to synchronize its
entire cache with that of the cache of its DCSs within a server
group. The CA message type code is 1. The CA message mandatory part
format is as follows:
Luciani, et al. [Page 24]
INTERNET-DRAFT SCSP Expires September 1997
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CA Sequence Number |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Mandatory Common Part |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CSAS Record |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
.......
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CSAS Record |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
CA Sequence Number
A value which provides a unique identifier to aid in the sequencing
of the cache alignment process. A "larger" sequence number means a
more recent CA message. The slave server always copies the
sequence number from the master server's previous CA message into
its current CA message which it is sending and the the slave
acknowledges the master's CA message. Since the initial CA process
is lock-step, if the slave does not receive the same sequence
number which it previously received then the information in the
slave's previous CA message is implicitly acknowledged. Note that
there is a separate CA Sequence Number space associated with each
CAFSM.
Mandatory Common Part
The mandatory common part is described in detail in Section
B.2.0.1. There are two fields in the mandatory common part whose
codings are specific to a given message type. These fields are the
"Number of Records" field and the "Flags" field.
Number of Records
The Number of Records field of the mandatory common part for the
CA message gives the number of CSAS Records appended to the CA
message mandatory part.
Flags
The Flags field of the mandatory common part for the CA message
has the following format:
0 1
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|M|I|O| unused |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Luciani, et al. [Page 25]
INTERNET-DRAFT SCSP Expires September 1997
M
This bit is part of the negotiation process for the cache
alignment. When this bit is set then the sender of the CA
message is indicating that it wishes to lead the alignment
process. This bit is the "Master/Slave bit".
I
When set, this bit indicates that the sender of the CA message
believes that it is in a state where it is negotiating for the
status of master or slave. This bit is the "Initialization
bit".
O
This bit indicates that the sender of the CA message has more
CSAS records to send. This implies that the cache alignment
process must continue. This bit is the "mOre bit" despite its
dubious name.
All other fields of the mandatory common part are coded as
described in Section B.2.0.1.
CSAS record
The CA message appends CSAS records to the end of its mandatory
part. These CSAS records are NOT embedded in CSA records. See
Section B.2.0.2 for details on CSAS records.
B.2.2 Cache State Update Request (CSU Request)
The Cache State Update Request (CSU Request) message is used to
update the state of cache entries in servers which are directly
connected to the server sending the message. A CSU Request message
is sent from one server (the LS) to directly connected server (the
DCS) when the LS observes changes in the state of one or more cache
entries. An LS observes such a change in state by either receiving a
CSU request which causes an update to the LS's database or by
observing a change of state of a cached entry originated by the LS.
The change in state of a cache entry is noted in a CSU message by
appending a "Cache State Advertisement" (CSA) record to the end of
the mandatory part of the CSU Request as shown below.
The CSU Request message type code is 2. The CSU Request message
mandatory part format is as follows:
Luciani, et al. [Page 26]
INTERNET-DRAFT SCSP Expires September 1997
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Mandatory Common Part |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CSA Record |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
.......
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CSA Record |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Mandatory Common Part
The mandatory common part is described in detail in Section
B.2.0.1. There are two fields in the mandatory common part whose
codings are specific to a given message type. These fields are the
"Number of Records" field and the "Flags" field.
Number of Records
The Number of Records field of the mandatory common part for the
CSU Request message gives the number of CSA Records appended to
the CSU Request message mandatory part.
Flags
Currently, there are no flags defined for the Flags field of the
mandatory common part for the CSU Request message.
All other fields of the mandatory common part are coded as
described in Section B.2.0.1.
CSA Record
See Section B.2.2.1.
B.2.2.1 Cache State Advertisement Record (CSA record)
CSA records contain the information necessary to relate the current
state of a cache entry in an SG to the servers being synchronized.
CSA records contain a CSAS Record header and a client/server protocol
specific part. The CSAS Record includes enough information for SCSP
to look into the client/server database for the appropriate database
cache entry and then compare the "newness" of the summary against the
"newness" of the cached entry. If the information contained in the
CSA is more new than the cached entry of the receiving server then
the cached entry is updated accordingly with the contents of the CSA
Record. The client/server protocol specific part of the CSA Record
is documented separately for each such protocol. Examples of the
protocol specific parts for NHRP and ATMARP are shown in [8] and [9]
Luciani, et al. [Page 27]
INTERNET-DRAFT SCSP Expires September 1997
respectively.
The amount of information carried by a specific CSA record may exceed
the size of a link layer PDU. Hence, such CSA records MUST be
fragmented across a number of CSU Request messages. The method by
which this is done, is client/server protocol specific and is
documented in the appropriate protocol specific document.
The content of a CSA record is as follows:
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| CSAS Record |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Client/Server Protocol Specific Part for cache entry ... |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
CSAS Record
See Section B.2.0.2 for rules and format for filling out a CSAS
Record when it is "embedded" in a CSA Record.
Client/Server Protocol Specific Part for cache entry
This field contains the fields which are specific to the protocol
specific portion of SCSP processing. The particular set of fields
are defined in separate documents for each protocol user of SCSP.
The Protocol ID, which identifies which protocol is using SCSP in
the given packet, is located in the mandatory part of the message.
B.2.3 Cache State Update Reply (CSU Reply)
The Cache State Update Reply (CSU Reply) message is sent from a DCS
to an LS to acknowledge one or more CSA records which were received
in a CSU Request. Reception of a CSA record in a CSU Request is
acknowledged by including a CSAS record in the CSU Reply which
corresponds to the CSA record being acknowledged. The CSU Reply
message is the same in format as the CSU Request message except for
the following: the type code is 3, only CSAS Records (rather than CSA
records) are returned, and only those CSAS Records for which CSA
Records are being acknowledged are returned. This implies that a
given LS sending a CSU Request may not receive an acknowledgment in a
single CSU Reply for all the CSA Records included in the CSU Request.
Luciani, et al. [Page 28]
INTERNET-DRAFT SCSP Expires September 1997
B.2.4 Cache State Update Solicit Message (CSUS message)
This message allows one server (LS) to solicit the entirety of CSA
record data stored in the cache of a directly connected server (DCS).
The DCS responds with CSU Request messages containing the appropriate
CSA records. The CSUS message type code is 4. The CSUS message
format is the same as that of the CSU Reply message. CSUS messages
solicit CSU Requests from only one server (the one identified by the
Receiver ID in the Mandatory Part of the message).
B.2.5 Hello:
The Hello message is used to check connectivity between the sending
server (the LS) and one of its directly connected neighbor servers
(the DCSs). The Hello message type code is 5. The Hello message
mandatory part format is as follows:
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| HelloInterval | DeadFactor |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| unused | Family ID |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Mandatory Common Part |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Additional Receiver ID Record |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
.........
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Additional Receiver ID Record |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
HelloInterval
The hello interval advertises the time between sending of
consecutive Hello Messages. If the LS does not receive a Hello
message from the DCS (which contains the LSID as a Receiver ID)
within the HelloInterval advertised by the DCS then the DCS's Hello
is considered to be late. Also, the LS MUST send its own Hello
message to a DCS within the HelloInterval which it advertised to
the DCS in the LS's previous Hello message to that DCS (otherwise
the DCS would consider the LS's Hello to be late).
DeadFactor
This is a multiplier to the HelloInterval. If an LS does not
receive a Hello message which contains the LS's LSID as a Receiver
ID within the interval HelloInterval*DeadFactor from a given DCS,
which advertised the HelloInterval and DeadFactor in a previous
Luciani, et al. [Page 29]
INTERNET-DRAFT SCSP Expires September 1997
Hello message, then the LS MUST consider the DCS to be stalled; at
this point, one of two things MUST happen: 1) if the LS has
received any Hello messages from the DCS during this time then the
LS transitions the corresponding HFSM to the Unidirectional State;
otherwise, 2) the LS transitions the corresponding HFSM to the
Waiting State.
Family ID
This is an opaque bit string which is used to refer to an aggregate
of Protocol ID/SGID pairs. Only a single HFSM is run for all
Protocol ID/SGID pairs assigned to a Family ID. Thus, there is a
one to many mapping between the single HFSM and the CAFSMs
corresponding to each of the Protocol ID/SGID pairs. This might
have the net effect of substantially reducing HFSM maintenance
traffic. See the protocol specific SCSP documents for further
details.
Mandatory Common Part
The mandatory common part is described in detail in Section
B.2.0.1. There are two fields in the mandatory common part whose
codings are specific to a given message type. These fields are the
"Number of Records" field and the "Flags" field.
Number of Records
The Number of Records field of the mandatory common part for the
Hello message contains the number of "Additional Receiver ID"
records which are included in the Hello. Additional Receiver ID
records contain a length field and a Receiver ID field. Note
that the count in "Number of Records" does NOT include the
Receiver ID which is included in the Mandatory Common Part.
Flags
Currently, there are no flags defined for the Flags field of the
mandatory common part for the Hello message.
All other fields of the mandatory common part are coded as
described in Section B.2.0.1.
Additional Receiver ID Record
This record contains a length field followed by a Receiver ID.
Since it is conceivable that the length of a given Receiver ID may
vary even within an SG, each additional Receiver ID heard (beyond
the first one) will have both its length in bytes and value encoded
in an "Additional Receiver ID Record". Receiver IDs are IDs of a
DCS from which the LS has heard a recent Hello (i.e., within
DeadFactor*HelloInterval as advertised by the DCS in a previous
Hello message).
Luciani, et al. [Page 30]
INTERNET-DRAFT SCSP Expires September 1997
The format for this record is as follows:
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Rec ID Len | Receiver ID |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
If the LS has not heard from any DCS then the LS sets the Hello
message fields as follows: Recvr ID Len is set to zero and no storage
is allocated for the Receiver ID in the Common Mandatory Part,
"Number of Records" is set to zero, and no storage is allocated for
"Additional Receiver ID Records".
If the LS has heard from exactly one DCS then the LS sets the Hello
message fields as follows: the Receiver ID of the DCS which was heard
and the length of that Receiver ID are encoded in the Common
Mandatory Part, "Number of Records" is set to zero, and no storage is
allocated for "Additional Receiver ID Records".
If the LS has heard from two or more DCSs then the LS sets the Hello
message fields as follows: the Receiver ID of the first DCS which was
heard and the length of that Receiver ID are encoded in the Common
Mandatory Part, "Number of Records" is set to the number of
"Additional" DCSs heard, and for each additional DCS an "Additional
Receiver ID Record" is formed and appended to the end of the Hello
message.
B.3 Extensions Part
The Extensions Part, if present, carries one or more extensions in
{Type, Length, Value} triplets.
Extensions have the following 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Value... |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Type
The extension type code (see below).
Luciani, et al. [Page 31]
INTERNET-DRAFT SCSP Expires September 1997
Length
The length in octets of the value (not including the Type and
Length fields; a null extension will have only an extension header
and a length of zero).
When extensions exist, the extensions part is terminated by the End
of Extensions extension, having Type = 0 and Length = 0.
Extensions may occur in any order but any particular extension type
may occur only once in an SCSP packet. An LS MUST NOT change the
order of extensions.
B.3.0 The End Of Extensions
Type = 0
Length = 0
When extensions exist, the extensions part is terminated by the End
Of Extensions extension.
B.3.1 SCSP Authentication Extension
Type = 1
Length = variable
The SCSP Authentication Extension is carried in SCSP packets to
convey authentication information between an LS and a DCS in the same
SG.
Authentication is done pairwise on an LS to DCS basis; i.e., the
authentication extension is generated at each LS. If a received
packet fails the authentication test then an "abnormal event" has
occurred. Any "abnormal event" causes the HFSM associated with the
server from which the packet was received to transition to the
Waiting State.
The presence or absence of the Authentication Extension is a local
matter.
Luciani, et al. [Page 32]
INTERNET-DRAFT SCSP Expires September 1997
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Authentication Type |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| |
+-+-+-+-+-+-+-+-+-+-+ Authentication Data... -+-+-+-+-+-+-+-+-+-+
| |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
The Authentication Type field identifies the authentication method in
use. Currently assigned values are:
1 - Cleartext Password
2 - Keyed MD5
All other values are reserved.
The Authentication Data field contains the type-specific
authentication information.
In the case of Cleartext Password Authentication, the Authentication
Data consists of a variable length password.
In the case of Keyed MD5 Authentication, the Authentication Data
contains the 16 byte MD5 digest of the entire SCSP packet, including
the encapsulated protocol's header, with the authentication key
appended to the end of the packet. The authentication key is not
transmitted with the packet. The MD5 digest covers only the fixed
part and mandatory part.
B.3.2 SCSP Vendor-Private Extension
Type = 2
Length = variable
The SCSP Vendor-Private Extension is carried in SCSP packets to
convey vendor-private information between an LS and a DCS in the same
SG and is thus of limited use. If a finer granularity (e.g., CSA
record level) is desired then then given client/server protocol
specific SCSP document MUST define such a mechanism. Obviously,
however, such a protocol specific mechanism might look exactly like
this extension. The Vendor Private Extension MAY NOT appear more
than once in an SCSP packet for a given Vendor ID value.
Luciani, et al. [Page 33]
INTERNET-DRAFT SCSP Expires September 1997
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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Vendor ID | Data.... |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Vendor ID
802 Vendor ID as assigned by the IEEE [10].
Data
The remaining octets after the Vendor ID in the payload are
vendor-dependent data.
If the receiver does not handle this extension, or does not match the
Vendor ID in the extension then the extension may be completely
ignored by the receiver.
References
[1] "Classical IP and ARP over ATM", Laubach, RFC 1577.
[2] "NBMA Next Hop Resolution Protocol (NHRP)", Luciani, Katz, Piscitello,
Cole, draft-ietf-rolc-nhrp-11.txt.
[3] "OSPF Version 2", Moy, RFC1583.
[4] "P-NNI V1", Dykeman, Goguen, 1996.
[5] "Support for Multicast over UNI 3.0/3.1 based ATM Networks.",
Armitage, RFC2022.
[6] LAN Emulation over ATM Version 2 - LNNI specification, Keene,
btd-lane-lnni-02.08.
[7] Assigned Numbers, Reynolds, Postel, RFC 1700.
[8] "A Distributed NHRP Service Using SCSP", Luciani,
draft-ietf-ion-scsp-nhrp-00.txt.
[9] "A Distributed ATMARP Service Using SCSP", Luciani,
draft-ietf-ion-scsp-atmarp-00.txt.
[10] Assigned Numbers, J. Reynolds and J. Postel, RFC 1700.
Luciani, et al. [Page 34]
INTERNET-DRAFT SCSP Expires September 1997
Acknowledgments
This I-D is a distillation of issues raised during private
discussions, on the IP-ATM mailing list, and during the Dallas IETF
(12/95). Thanks to all who have contributed but particular thanks to
following people (in no particular order): Barbara Fox of Harris and
Jeffries; Raj Nair, and Matthew Doar of Ascom Nexion; Andy Malis of
Cascade; Andre Fredette of Bay Networks; James Watt of Newbridge; and
Carl Marcinik of Fore.
Author's Address
James V. Luciani
Bay Networks, Inc.
3 Federal Street, BL3-04
Billerica, MA 01821
phone: +1-508-916-4734
email: luciani@baynetworks.com
Grenville Armitage
Bellcore, 445 South Street
Morristown, NJ, 07960
Email: gja@thumper.bellcore.com
Ph. +1 201 829 2635
Joel M. Halpern
Newbridge Networks Corp.
593 Herndon Parkway
Herndon, VA 22070-5241
Phone: +1-703-708-5954
Email: jhalpern@Newbridge.COM
Luciani, et al. [Page 35]