draft-ietf-sigtran-sctp-05.txt   draft-ietf-sigtran-sctp-06.txt 
skipping to change at page 10, line ? skipping to change at page 91, line ?
Nortel Networks Nortel Networks
I. Rytina I. Rytina
Ericsson Ericsson
M. Kalla M. Kalla
Telcordia Telcordia
L. Zhang L. Zhang
UCLA UCLA
V. Paxson V. Paxson
ACIRI ACIRI
expires in six months January 17,2000 expires in six months February 23,2000
Simple Control Transmission Protocol Simple Control Transmission Protocol
<draft-ietf-sigtran-sctp-05.txt> <draft-ietf-sigtran-sctp-06.txt>
Status of This Memo Status of This Memo
This document is an Internet-Draft and is in full conformance with all This document is an Internet-Draft and is in full conformance with all
provisions of Section 10 of RFC 2026. Internet-Drafts are working provisions of Section 10 of RFC 2026. Internet-Drafts are working
documents of the Internet Engineering Task Force (IETF), its areas, documents of the Internet Engineering Task Force (IETF), its areas,
and its working groups. Note that other groups may also distribute and its working groups. Note that other groups may also distribute
working documents as Internet-Drafts. working documents as Internet-Drafts.
The list of current Internet-Drafts can be accessed at The list of current Internet-Drafts can be accessed at
http://www.ietf.org/ietf/1id-abstracts.txt http://www.ietf.org/ietf/1id-abstracts.txt
The list of Internet-Draft Shadow Directories can be accessed at The list of Internet-Draft Shadow Directories can be accessed at
http://www.ietf.org/shadow.html. http://www.ietf.org/shadow.html.
Stewart, et al [Page 1] Stewart, et al [Page 1]
Abstract Abstract
This document describes the Simple Control Transmission Protocol This document describes the Simple Control Transmission Protocol
(SCTP). SCTP is designed to transport PSTN signalling messages over (SCTP). SCTP is designed to transport PSTN signaling messages over
IP networks, but is capable of broader applications. IP networks, but is capable of broader applications.
SCTP is a reliable datagram transfer protocol operating on top of an SCTP is a reliable datagram transfer protocol operating on top of an
unreliable routed packet network such as IP. It offers the following unreliable routed packet network such as IP. It offers the following
services to its users: services to its users:
-- acknowledged error-free non-duplicated transfer of user data, -- acknowledged error-free non-duplicated transfer of user data,
-- data segmentation to conform to discovered path MTU size, -- data segmentation to conform to discovered path MTU size,
-- sequenced delivery of user messages within multiple streams, -- sequenced delivery of user messages within multiple streams,
with an option for order-of-arrival delivery of individual with an option for order-of-arrival delivery of individual
user messages, user messages,
-- optional multiplexing of user messages into SCTP datagrams, and -- optional multiplexing of user messages into SCTP datagrams, and
-- network-level fault tolerance through supporting of multi-homing -- network-level fault tolerance through supporting of multi-homing
at either or both ends of an association. at either or both ends of an association.
The design of SCTP includes appropriate congestion avoidance behaviour The design of SCTP includes appropriate congestion avoidance behavior
and resistance to flooding and masquerade attacks. and resistance to flooding and masquerade attacks.
Stewart, et al [Page 2] Stewart, et al [Page 2]
TABLE OF CONTENTS TABLE OF CONTENTS
1. Introduction.................................................. 5 1. Introduction.................................................. 5
1.1 Motivation.................................................. 5 1.1 Motivation.................................................. 5
1.2 Architectural View of SCTP.................................. 6 1.2 Architectural View of SCTP..................................5
1.3 Functional View of SCTP..................................... 6 1.3 Functional View of SCTP..................................... 6
1.3.1 Association Startup and Takedown........................ 7 1.3.1 Association Startup and Takedown........................ 7
1.3.2 Sequenced Delivery within Streams....................... 8 1.3.2 Sequenced Delivery within Streams.......................7
1.3.3 User Data Segmentation.................................. 8 1.3.3 User Data Segmentation.................................. 8
1.3.4 Acknowledgement and Congestion Avoidance................ 8 1.3.4 Acknowledgment and Congestion Avoidance.................8
1.3.5 Chunk Multiplex......................................... 9 1.3.5 Chunk Multiplex.........................................8
1.3.6 Path Management......................................... 9 1.3.6 Path Management.........................................8
1.3.7 Message Validation...................................... 9 1.3.7 Message Validation...................................... 9
1.4 Recapitulation of Key Terms.................................10 1.4 Recapitulation of Key Terms.................................9
1.5 Abbreviations...............................................12 1.5 Abbreviations...............................................11
2. SCTP Datagram Format..........................................12 2. SCTP Datagram Format..........................................11
2.1 SCTP Common Header Field Descriptions.......................13 2.1 SCTP Common Header Field Descriptions.......................12
2.2 Chunk Field Descriptions....................................14 2.2 Chunk Field Descriptions....................................13
2.2.1 Optional/Variable-length Parameter Format...............16 2.2.1 Optional/Variable-length Parameter Format...............14
2.2.2 Vendor-Specific Extension Parameter Format..............16 2.2.2 Vendor-Specific Extension Parameter Format..............15
2.3 SCTP Chunk Definitions......................................18 2.3 SCTP Chunk Definitions......................................17
2.3.1 Initiation (INIT).......................................18 2.3.1 Initiation (INIT).......................................17
2.3.1.1 Optional or Variable Length Parameters..............20 2.3.1.1 Optional or Variable Length Parameters..............19
2.3.2 Initiation Acknowledgement (INIT ACK)...................23 2.3.2 Initiation Acknowledgment (INIT ACK)....................20
2.3.2.1 Optional or Variable Length Parameters..............24 2.3.2.1 Optional or Variable Length Parameters..............21
2.3.3 Selective Acknowledgement (SACK)........................25 2.3.3 Selective Acknowledgment (SACK).........................22
2.3.4 Heartbeat Request (HEARTBEAT)...........................27 2.3.4 Heartbeat Request (HEARTBEAT)...........................25
2.3.5 Heartbeat Acknowledgment (HEARTBEAT ACK)................28 2.3.5 Heartbeat Acknowledgment (HEARTBEAT ACK)................26
2.3.6 Abort Association (ABORT)...............................29 2.3.6 Abort Association (ABORT)...............................26
2.3.7 Shutdown Association (SHUTDOWN).........................30 2.3.7 Shutdown Association (SHUTDOWN).........................27
2.3.8 Shutdown Acknowledgment (SHUTDOWN ACK)..................30 2.3.8 Shutdown Acknowledgment (SHUTDOWN ACK)..................28
2.3.9 Operation Error (ERROR).................................31 2.3.9 Operation Error (ERROR).................................28
2.3.10 Encryption Cookie (COOKIE).............................33 2.3.10 Encryption Cookie (COOKIE).............................30
2.3.11 Cookie Acknowledgment (COOKIE ACK).....................33 2.3.11 Cookie Acknowledgment (COOKIE ACK).....................31
2.3.12 Payload Data (DATA)....................................34 2.3.12 Payload Data (DATA)....................................31
2.4 Vendor-Specific Chunk Extensions............................35 2.4 Vendor-Specific Chunk Extensions............................33
3. SCTP Association State Diagram.................................37 3. SCTP Association State Diagram.................................34
4. Association Initialization.....................................39 4. Association Initialization.....................................36
4.1 Normal Establishment of an Association......................39 4.1 Normal Establishment of an Association......................37
4.1.1 Handle Stream Parameters................................41 4.1.1 Handle Stream Parameters................................38
4.1.2 Handle Address Parameters...............................41 4.1.2 Handle Address Parameters...............................39
4.1.3 Generating Encryption Cookie............................41 4.1.3 Generating Encryption Cookie............................39
4.1.4 Cookie Processing.......................................42 4.1.4 Cookie Processing.......................................40
4.1.5 Cookie Authentication...................................42 4.1.5 Cookie Authentication...................................40
4.1.6 An Example of Normal Association Establishment..........43 4.1.6 An Example of Normal Association Establishment..........41
4.2 Handle Duplicate INIT, INIT ACK, COOKIE, and COOKIE ACK.....44 4.2 Handle Duplicate INIT, INIT ACK, COOKIE, and COOKIE ACK.....42
4.2.1 Handle Duplicate INIT in COOKIE-WAIT 4.2.1 Handle Duplicate INIT in COOKIE-WAIT
or COOKIE-SENT States...................................45 or COOKIE-SENT States...................................43
4.2.2 Handle Duplicate INIT in Other States...................45 4.2.2 Handle Duplicate INIT in Other States...................43
4.2.3 Handle Duplicate INIT ACK...............................46 4.2.3 Handle Duplicate INIT ACK...............................43
4.2.4 Handle Duplicate COOKIE.................................46 4.2.4 Handle Duplicate COOKIE.................................43
4.2.5 Handle Duplicate COOKIE-ACK.............................47 4.2.5 Handle Duplicate COOKIE-ACK.............................45
4.2.6 Handle Stale COOKIE Error...............................45
4.3 Other Initialization Issues.................................45
Stewart, et al [Page 3] Stewart, et al [Page 3]
4.2.6 Handle Stale COOKIE Error...............................47 4.3.1 Selection of Tag Value..................................45
4.3 Other Initialization Issues.................................48 5. User Data Transfer.............................................46
4.3.1 Selection of Tag Value..................................48 5.1 Transmission of DATA Chunks.................................47
5. User Data Transfer.............................................48 5.2 Acknowledgment of Reception of DATA Chunks..................48
5.1 Transmission of DATA Chunks.................................49 5.2.1 Tracking Peer's Receive Buffer Space....................49
5.2 Acknowledgment of Reception of DATA Chunks..................51 5.3 Management Retransmission Timer.............................49
5.2.1 Tracking Peer's Receive Buffer Space....................52 5.3.1 RTO Calculation.........................................50
5.3 Management Retransmission Timer.............................52 5.3.2 Retransmission Timer Rules..............................51
5.3.1 RTO Calculation.........................................52 5.3.3 Handle T3-rxt Expiration................................52
5.3.2 Retransmission Timer Rules..............................53 5.4 Multi-homed SCTP Endpoints..................................52
5.3.3 Handle T3-rxt Expiration................................54 5.4.1 Failover from Inactive Destination Address..............53
5.4 Multi-homed SCTP Endpoints..................................55 5.5 Stream Identifier and Stream Sequence Number................53
5.4.1 Failover from Inactive Destination Address..............56 5.6 Ordered and Un-ordered Delivery.............................54
5.5 Stream Identifier and Sequence Number.......................56 5.7 Report Gaps in Received DATA TSNs...........................54
5.6 Ordered and Un-ordered Delivery.............................56 5.8 Adler-32 Checksum Calculation...............................55
5.7 Report Gaps in Received DATA TSNs...........................57 5.9 Segmentation................................................56
5.8 Adler-32 Checksum Calculation...............................58 5.10 Bundling and Multiplexing..................................57
5.9 Segmentation................................................59 6. Congestion Control ..........................................57
5.10 Bundling and Multiplexing..................................60 6.1 SCTP Differences from TCP Congestion Control................58
6. Congestion Control ..........................................60 6.2 SCTP Slow-Start and Congestion Avoidance....................59
6.1 SCTP Differences from TCP Congestion Control................61 6.2.1 Slow-Start..............................................59
6.2 SCTP Slow-Start and Congestion Avoidance....................62 6.2.2 Congestion Avoidance....................................60
6.2.1 Slow-Start..............................................62 6.2.3 Congestion Control......................................61
6.2.2 Congestion Avoidance....................................63 6.2.4 Fast Retransmit on Gap Reports..........................61
6.2.3 Congestion Control......................................63 6.3 Path MTU Discovery..........................................62
6.2.4 Fast Retransmit on Gap Reports..........................64 7. Fault Management..............................................63
6.3 Path MTU Discovery..........................................64 7.1 Endpoint Failure Detection..................................63
7. Fault Management..............................................65 7.2 Path Failure Detection......................................63
7.1 Endpoint Failure Detection..................................65 7.3 Path Heartbeat..............................................64
7.2 Path Failure Detection......................................66 7.4 Handle "Out of the blue" Packets............................65
7.3 Path Heartbeat..............................................66 7.5 Verification Tag............................................65
7.4 Handle "Out of the blue" Packets............................67 7.5.1 Exceptions in Verification Tag Rules....................66
7.5 Verification Tag............................................67 8. Termination of Association.....................................66
7.5.1 Exceptions in Verification Tag Rules....................67 8.1 Close of an Association.....................................66
8. Termination of Association.....................................68 8.2 Shutdown of an Association..................................67
8.1 Close of an Association.....................................68 9. Interface with Upper Layer.....................................68
8.2 Shutdown of an Association..................................68 9.1 ULP-to-SCTP.................................................68
9. Interface with Upper Layer.....................................69 9.2 SCTP-to-ULP.................................................76
9.1 ULP-to-SCTP.................................................70 10. Security Considerations.......................................79
9.2 SCTP-to-ULP.................................................77 10.1 Security Objectives........................................79
10. Security Considerations.......................................80 10.2 SCTP Responses To Potential Threats........................79
10.1 Security Objectives........................................80 10.2.1 Countering Insider Attacks.............................79
10.2 SCTP Responses To Potential Threats........................80 10.2.2 Protecting against Data Corruption in the Network......79
10.2.1 Countering Insider Attacks.............................80 10.2.3 Protecting Confidentiality.............................80
10.2.2 Protecting against Data Corruption in the Network......80 10.2.4 Protecting against Blind Denial of Service Attacks.....80
10.2.3 Protecting Confidentiality.............................81 10.2.4.1 Flooding...........................................80
10.2.4 Protecting against Blind Denial of Service Attacks.....81 10.2.4.2 Masquerade.........................................81
10.2.4.1 Flooding...........................................81 10.2.4.3 Improper Monopolization of Services................81
10.2.4.2 Masquerade.........................................82 10.3 Protection against Fraud and Repudiation...................82
10.2.4.3 Improper Monopolization of Services................83 11. Recommended Transmission Control Block (TCB) Parameters.......83
10.3 Protection against Fraud and Repudiation...................83 12. IANA Consideration............................................86
11. IANA Consideration............................................84 12.1 IETF-defined Chunk Extension...............................86
11.1 IETF-defined Chunk Extension...............................84 12.2 IETF-defined Chunk Parameter Extension.....................87
12.3 IETF-defined Additional Error Causes.......................88
12.4 Payload Protocol Identifiers...............................88
Stewart, et al [Page 4] Stewart, et al [Page 4]
11.2 IETF-defined Chunk Parameter Extension.....................85
11.3 IETF-defined Additional Error Causes.......................85 13. Suggested SCTP Protocol Parameter Values......................89
11.4 Payload Protocol Identifiers...............................86 14. Acknowledgments...............................................89
12. Suggested SCTP Protocol Parameter Values......................86 15. Authors' Addresses............................................89
13. Acknowledgments...............................................87 16. References....................................................90
14. Authors' Addresses............................................87
15. References....................................................88
1. Introduction 1. Introduction
This section explains the reasoning behind the development of the This section explains the reasoning behind the development of the
Simple Control Transmission Protocol (SCTP), the services it offers, Simple Control Transmission Protocol (SCTP), the services it offers,
and the basic concepts needed to understand the detailed description and the basic concepts needed to understand the detailed description
of the protocol. of the protocol.
1.1 Motivation 1.1 Motivation
TCP [8] has performed immense service as the primary means of reliable TCP [8] has performed immense service as the primary means of reliable
data transfer in IP networks. However, an increasing number of recent data transfer in IP networks. However, an increasing number of recent
applications have found TCP too limiting, and have incorporated their applications have found TCP too limiting, and have incorporated their
own reliable data transfer protocol on top of UDP [9]. The limitations own reliable data transfer protocol on top of UDP [9]. The limitations
which users have wished to bypass relate both to the intrinsic nature which users have wished to bypass include the following:
of TCP and to its typical implementation.
Intrinsic limitations:
-- TCP provides both reliable data transfer and strict order- -- TCP provides both reliable data transfer and strict order-
of-transmission delivery of data. Some applications need reliable of-transmission delivery of data. Some applications need reliable
transfer without sequence maintenance, while others would be transfer without sequence maintenance, while others would be
satisfied with partial ordering of the data. In both of these satisfied with partial ordering of the data. In both of these
cases the head-of-line blocking offered by TCP causes cases the head-of-line blocking offered by TCP causes
unnecessary delay. unnecessary delay.
-- The stream-oriented nature of TCP is often an inconvenience. -- The stream-oriented nature of TCP is often an inconvenience.
Applications must add their own record marking to delineate Applications must add their own record marking to delineate
their messages, and must make explicit use of the push facility their messages, and must make explicit use of the push facility
to ensure that a complete message is transferred in a to ensure that a complete message is transferred in a
reasonable time. reasonable time.
-- The limited scope of TCP sockets complicates the task of -- The limited scope of TCP sockets complicates the task of
providing highly-available data transfer capability using providing highly-available data transfer capability using
multi-homed hosts. multi-homed hosts.
Limitations due to implementation: -- TCP is relatively vulnerable to denial of service attacks,
such as SYN attacks.
-- TCP is generally implemented at the operating system level.
Kernel limitations may constrain the maximum allowable number
Stewart, et al [Page 5]
of simultaneous TCP connections to a number far below that
required for certain applications.
-- TCP implementations do not generally allow the application
to control the timers which determine how quickly a connection
failure is discovered. Some applications are more critically
dependent than others on timely initiation of recovery from
such failures.
Transport of PSTN signalling across the IP network is an application Transport of PSTN signaling across the IP network is an application
for which all of these limitations of TCP are relevant. While this for which all of these limitations of TCP are relevant. While this
application directly motivated the development of SCTP, other application directly motivated the development of SCTP, other
applications may find SCTP a good match to their requirements. applications may find SCTP a good match to their requirements.
1.2 Architectural View of SCTP 1.2 Architectural View of SCTP
SCTP is viewed as a layer between the SCTP user application ("SCTP SCTP is viewed as a layer between the SCTP user application ("SCTP
user" for short) and an unreliable routed packet network service such user" for short) and an unreliable routed packet network service such
as IP. The basic service offered by SCTP is the reliable transfer of as IP. The basic service offered by SCTP is the reliable transfer of
user messages between peer SCTP users. It performs this service user messages between peer SCTP users. It performs this service
Stewart, et al [Page 5]
within the context of an association between two SCTP nodes. Chapter 9 within the context of an association between two SCTP nodes. Chapter 9
of this document sketches the API which should exist at the boundary of this document sketches the API which should exist at the boundary
between the SCTP and the SCTP user layers. between the SCTP and the SCTP user layers.
SCTP is connection-oriented in nature, but the SCTP association is a SCTP is connection-oriented in nature, but the SCTP association is a
broader concept than the TCP connection. SCTP provides the means for broader concept than the TCP connection. SCTP provides the means for
each SCTP endpoint (Section 1.4) to provide the other during each SCTP endpoint (Section 1.4) to provide the other during
association startup with a list of transport addresses (e.g. multiple association startup with a list of transport addresses (e.g. multiple
IP addresses in combination with an SCTP port) through which that IP addresses in combination with an SCTP port) through which that
endpoint can be reached and from which it will originate messages. endpoint can be reached and from which it will originate messages.
skipping to change at page 10, line ? skipping to change at page 91, line ?
|-------------| |-------------| |-------------| |-------------|
| |One or more ---- One or more| | | |One or more ---- One or more| |
| IP Network |IP address \/ IP address| IP Network | | IP Network |IP address \/ IP address| IP Network |
| Service |appearances /\ appearances| Service | | Service |appearances /\ appearances| Service |
|_____________| ---- |_____________| |_____________| ---- |_____________|
SCTP Node A |<-------- Network transport ------->| SCTP Node B SCTP Node A |<-------- Network transport ------->| SCTP Node B
Figure 1: An SCTP Association Figure 1: An SCTP Association
Stewart, et al [Page 6]
1.3 Functional View of SCTP 1.3 Functional View of SCTP
The SCTP transport service can be decomposed into a number of The SCTP transport service can be decomposed into a number of
functions. These are depicted in Figure 2 and explained in the functions. These are depicted in Figure 2 and explained in the
remainder of this section. remainder of this section.
SCTP User Application SCTP User Application
..----------------------------------------------------- ..-----------------------------------------------------
.. _____________ ____________________ .. _____________ ____________________
| | | Sequenced delivery | | | | Sequenced delivery |
| Association | | within streams | | Association | | within streams |
| | |____________________| | | |____________________|
| startup | | startup |
..| | ____________________________ ..| | ____________________________
| and | | User Data Segmentation | | and | | User Data Segmentation |
| | |____________________________| | | |____________________________|
| takedown | | takedown |
Stewart, et al [Page 6]
..| | ____________________________ ..| | ____________________________
| | | Acknowledgement | | | | Acknowledgment |
| | | and | | | | and |
| | | Congestion Avoidance | | | | Congestion Avoidance |
..| | |____________________________| ..| | |____________________________|
| | | |
| | ____________________________ | | ____________________________
| | | Chunk Multiplex | | | | Chunk Multiplex |
| | |____________________________| | | |____________________________|
| | | |
| | ________________________________ | | ________________________________
| | | Path Management | | | | Message Validataion |
| | |________________________________| | | |________________________________|
| | | |
| | ________________________________ | | ________________________________
| | | Message Validation | | | | Path Management |
|______________ |________________________________| |______________ |________________________________|
Figure 2: Functional View of the SCTP Transport Service Figure 2: Functional View of the SCTP Transport Service
1.3.1 Association Startup and Takedown 1.3.1 Association Startup and Takedown
An association is initiated by a request from the SCTP user (see the An association is initiated by a request from the SCTP user (see the
description of the ASSOCIATE primitive in Chapter 9). description of the ASSOCIATE primitive in Chapter 9).
A cookie mechanism, taken from that devised by Karn and Simpson in RFC A cookie mechanism, taken from that devised by Karn and Simpson in RFC
2522 [6], is employed during the initialization to provide protection 2522 [6], is employed during the initialization to provide protection
against security attacks. The cookie mechanism uses a four-way against security attacks. The cookie mechanism uses a four-way
handshaking, but the last two legs of which are allowed to carry user handshaking, but the last two legs of which are allowed to carry user
Stewart, et al [Page 7]
data for fast setup. The startup sequence is described in chapter 4 of data for fast setup. The startup sequence is described in chapter 4 of
this document. this document.
SCTP provides for graceful takedown of an active association on SCTP provides for graceful takedown of an active association on
request from the SCTP user. See the description of the TERMINATE request from the SCTP user. See the description of the TERMINATE
primitive in chapter 9. SCTP also allows ungraceful takedown, either primitive in chapter 9. SCTP also allows ungraceful takedown, either
on request from the user (ABORT primitive) or as a result of an error on request from the user (ABORT primitive) or as a result of an error
condition detected within the SCTP layer. Chapter 8 describes both the condition detected within the SCTP layer. Chapter 8 describes both the
graceful and the ungraceful takedown procedures. graceful and the ungraceful takedown procedures.
skipping to change at page 10, line ? skipping to change at page 91, line ?
The term "stream" is used in SCTP to refer to a sequence of user The term "stream" is used in SCTP to refer to a sequence of user
messages. This is in contrast to its usage in TCP, where it refers to messages. This is in contrast to its usage in TCP, where it refers to
a sequence of bytes. a sequence of bytes.
The SCTP user can specify at association startup time the number of The SCTP user can specify at association startup time the number of
streams to be supported by the association. This number is negotiated streams to be supported by the association. This number is negotiated
with the remote end (see section 4.1.1). User messages are associated with the remote end (see section 4.1.1). User messages are associated
with stream numbers (SEND, RECEIVE primitives, Chapter 9). Internally, with stream numbers (SEND, RECEIVE primitives, Chapter 9). Internally,
SCTP assigns a stream sequence number to each message passed to it by SCTP assigns a stream sequence number to each message passed to it by
Stewart, et al [Page 7]
the SCTP user. On the receiving side, SCTP ensures that messages are the SCTP user. On the receiving side, SCTP ensures that messages are
delivered to the SCTP user in sequence within a given stream. However, delivered to the SCTP user in sequence within a given stream. However,
while one stream may be blocked waiting for the next in-sequence user while one stream may be blocked waiting for the next in-sequence user
message, delivery from other streams may proceed. message, delivery from other streams may proceed.
SCTP provides a mechanism for bypassing the sequenced delivery SCTP provides a mechanism for bypassing the sequenced delivery
service. User messages sent using this mechanism are delivered to the service. User messages sent using this mechanism are delivered to the
SCTP user as soon as they are received. SCTP user as soon as they are received.
1.3.3 User Data Segmentation 1.3.3 User Data Segmentation
SCTP can segment user messages to ensure that the SCTP datagram SCTP can segment user messages to ensure that the SCTP datagram
passed to the lower layer conforms to the path MTU. Segments are passed to the lower layer conforms to the path MTU. Segments are
reassembled into complete messages before being passed to the SCTP reassembled into complete messages before being passed to the SCTP
user. user.
1.3.4 Acknowledgement and Congestion Avoidance 1.3.4 Acknowledgment and Congestion Avoidance
SCTP assigns a Transmission Sequence Number (TSN) to each user data SCTP assigns a Transmission Sequence Number (TSN) to each user data
segment or unsegmented message. The TSN is independent of any segment or unsegmented message. The TSN is independent of any
sequence number assigned at the stream level. The receiving end stream sequence number assigned at the stream level. The receiving end
acknowledges all TSNs received, even if there are gaps in the acknowledges all TSNs received, even if there are gaps in the
sequence. In this way, reliable delivery is kept functionally separate sequence. In this way, reliable delivery is kept functionally separate
from sequenced delivery. from sequenced delivery.
The Acknowledgement and Congestion Avoidance function is responsible The Acknowledgment and Congestion Avoidance function is responsible
for message retransmission when timely acknowledgement has not been for message retransmission when timely acknowledgment has not been
Stewart, et al [Page 8]
received. Message retransmission is conditioned by congestion received. Message retransmission is conditioned by congestion
avoidance procedures similar to those used for TCP. See Chapters 5 avoidance procedures similar to those used for TCP. See Chapters 5
and 6 for a detailed description of the protocol procedures associated and 6 for a detailed description of the protocol procedures associated
with this function. with this function.
1.3.5 Chunk Multiplex 1.3.5 Chunk Multiplex
As described in Chapter 2, the SCTP datagram as delivered to the lower As described in Chapter 2, the SCTP datagram as delivered to the lower
layer consists of a common header followed by one or more chunks. Each layer consists of a common header followed by one or more chunks. Each
chunk may contain either user data or SCTP control information. The chunk may contain either user data or SCTP control information. The
skipping to change at page 10, line ? skipping to change at page 91, line ?
1.3.6 Path Management 1.3.6 Path Management
The sending SCTP user is able to manipulate the set of transport The sending SCTP user is able to manipulate the set of transport
addresses used as destinations for SCTP datagrams, through the addresses used as destinations for SCTP datagrams, through the
primitives described in Chapter 9. The SCTP path management function primitives described in Chapter 9. The SCTP path management function
chooses the destination transport address for each outgoing SCTP chooses the destination transport address for each outgoing SCTP
datagram based on the SCTP user's instructions and the currently datagram based on the SCTP user's instructions and the currently
perceived reachability status of the eligible destination set. perceived reachability status of the eligible destination set.
Stewart, et al [Page 8]
The path management function monitors reachability through heartbeat The path management function monitors reachability through heartbeat
messages when other message traffic is inadequate to provide this messages when other message traffic is inadequate to provide this
information, and advises the SCTP user when reachability of any far- information, and advises the SCTP user when reachability of any far-
end transport address changes. The path management function is also end transport address changes. The path management function is also
responsible for reporting the eligible set of local transport responsible for reporting the eligible set of local transport
addresses to the far end during association startup, and for reporting addresses to the far end during association startup, and for reporting
the transport addresses returned from the far end to the SCTP user. the transport addresses returned from the far end to the SCTP user.
At association start-up, a primary destination transport address is At association start-up, a primary destination transport address is
defined for each SCTP endpoint, and is used for normal sending of SCTP defined for each SCTP endpoint, and is used for normal sending of SCTP
skipping to change at page 10, line ? skipping to change at page 91, line ?
On the receiving end, the path management is responsible for verifying On the receiving end, the path management is responsible for verifying
the existence of a valid SCTP association to which the inbound SCTP the existence of a valid SCTP association to which the inbound SCTP
datagram belongs before passing it for further processing. datagram belongs before passing it for further processing.
1.3.7 Message Validation 1.3.7 Message Validation
A mandatory verification tag and an Adler-32 checksum [2] fields are A mandatory verification tag and an Adler-32 checksum [2] fields are
included in the SCTP common header. The verification tag value is included in the SCTP common header. The verification tag value is
chosen by each end of the association during association startup. chosen by each end of the association during association startup.
Messages received without the verification tag value expected by the Messages received without the verification tag value expected by the
Stewart, et al [Page 9]
receiver are discarded, as a protection against blind masquerade receiver are discarded, as a protection against blind masquerade
attacks and against stale datagrams from a previous association. attacks and against stale datagrams from a previous association.
The Adler-32 checksum MUST be set by the sender of each SCTP datagram, The Adler-32 checksum should be set by the sender of each SCTP datagram,
to provide additional protection against data corruption in the to provide additional protection against data corruption in the
network beyond that provided by lower layers (e.g. the IP checksum). network beyond that provided by lower layers (e.g. the IP checksum).
1.4 Recapitulation of Key Terms 1.4 Recapitulation of Key Terms
The language used to describe SCTP has been introduced in the previous The language used to describe SCTP has been introduced in the previous
sections. This section provides a consolidated list of the key terms sections. This section provides a consolidated list of the key terms
and their definitions. and their definitions.
o SCTP user application (SCTP user): The logical higher-layer o SCTP user application (SCTP user): The logical higher-layer
skipping to change at page 10, line ? skipping to change at page 91, line ?
between SCTP and the unreliable packet network (e.g. IP) which between SCTP and the unreliable packet network (e.g. IP) which
it is using. An SCTP datagram includes the common SCTP header, it is using. An SCTP datagram includes the common SCTP header,
possible SCTP control chunks, and user data encapsulated within possible SCTP control chunks, and user data encapsulated within
SCTP DATA chunks. SCTP DATA chunks.
o Transport address: an address which serves as a source or o Transport address: an address which serves as a source or
destination for the unreliable packet transport service used by destination for the unreliable packet transport service used by
SCTP. In IP networks, a transport address is defined by the SCTP. In IP networks, a transport address is defined by the
combination of an IP address and an SCTP port number. combination of an IP address and an SCTP port number.
Stewart, et al [Page 9]
Note, only one SCTP port may be defined for each endpoint, Note, only one SCTP port may be defined for each endpoint,
but each endpoint may have multiple IP addresses. but each endpoint may have multiple IP addresses.
o SCTP endpoint: the logical sender/receiver of SCTP datagrams. On a o SCTP endpoint: the logical sender/receiver of SCTP datagrams. On a
multi-homed host, an SCTP endpoint is represented to its peers as a multi-homed host, an SCTP endpoint is represented to its peers as a
combination of a set of eligible destination transport addresses to combination of a set of eligible destination transport addresses to
which SCTP datagrams can be sent and a set of eligible source which SCTP datagrams can be sent and a set of eligible source
transport addresses from which SCTP datagrams can be received. transport addresses from which SCTP datagrams can be received.
Note, a source or destination transport address can only be Note, a source or destination transport address can only be
skipping to change at page 11, line 20 skipping to change at page 91, line ?
user data to permit the receiving SCTP endpoint to acknowledge its user data to permit the receiving SCTP endpoint to acknowledge its
receipt and detect duplicate deliveries. receipt and detect duplicate deliveries.
o Stream: a uni-directional logical channel established from one to o Stream: a uni-directional logical channel established from one to
another associated SCTP endpoints, within which all user messages another associated SCTP endpoints, within which all user messages
are delivered in sequence except for those submitted to the are delivered in sequence except for those submitted to the
un-ordered delivery service. un-ordered delivery service.
Note: The relationship between stream numbers in opposite Note: The relationship between stream numbers in opposite
directions is strictly a matter of how the applications use directions is strictly a matter of how the applications use
them. It is the responsiblity of the SCTP user to create and them. It is the responsibility of the SCTP user to create and
manage these correlations if they are so desired. manage these correlations if they are so desired.
o Stream Sequence Number: a 16-bit sequence number used internally by o Stream Sequence Number: a 16-bit sequence number used internally by
SCTP to assure sequenced delivery of the user messages within a SCTP to assure sequenced delivery of the user messages within a
given stream. One stream sequence number is attached to each user given stream. One stream sequence number is attached to each user
message. message.
o Path: the route taken by the SCTP datagrams sent by one SCTP o Path: the route taken by the SCTP datagrams sent by one SCTP
endpoint to a specific destination transport address of its peer endpoint to a specific destination transport address of its peer
SCTP endpoint. Note, sending to different destination transport SCTP endpoint. Note, sending to different destination transport
addresses does not necessarily guarantee getting separate paths. addresses does not necessarily guarantee getting separate paths.
o Bundling: an optional multiplexing operation, whereby more than one o Bundling: an optional multiplexing operation, whereby more than one
user messages may be carried in the same SCTP datagram. Each user user messages may be carried in the same SCTP datagram. Each user
message occupies its own DATA chunk. message occupies its own DATA chunk.
o Outstanding TSN (at an SCTP endpoint): a TSN (and the associated DATA o Outstanding TSN (at an SCTP endpoint): a TSN (and the associated DATA
chunk) which have been sent by the endpoint but for which it has not chunk) which have been sent by the endpoint but for which it has not
yet received an acknowledgement. yet received an acknowledgment.
Stewart, et al [Page 10]
o Unacknowledged TSN (at an SCTP endpoint): a TSN (and the associated DATA o Unacknowledged TSN (at an SCTP endpoint): a TSN (and the associated DATA
chunk) which have been received by the endpoint but for which an chunk) which have been received by the endpoint but for which an
acknowledgement has not yet been sent. acknowledgment has not yet been sent.
o Receiver Window (rwnd): The most recently calculated receiver o Receiver Window (rwnd): The most recently calculated receiver
window, in number of octets. This gives an indication of the space window, in number of octets. This gives an indication of the space
available in the receiver's inbound buffer. available in the receiver's inbound buffer.
o Congestion Window (cwnd): An SCTP variable that limits the data, in o Congestion Window (cwnd): An SCTP variable that limits the data, in
number of octets, a sender can send into the network before number of octets, a sender can send into the network before
receiving an acknowledgment on a particular destination Transport receiving an acknowledgment on a particular destination Transport
address. address.
skipping to change at page 12, line 37 skipping to change at page 91, line ?
TSN - Transmission Sequence Number TSN - Transmission Sequence Number
ULP - Upper-layer Protocol ULP - Upper-layer Protocol
2. SCTP Datagram Format 2. SCTP Datagram Format
An SCTP datagram is composed of a common header and chunks. A chunk An SCTP datagram is composed of a common header and chunks. A chunk
contains either control information or user data. contains either control information or user data.
Stewart, et al [Page 11]
The SCTP datagram format is shown below: The SCTP datagram format is shown below:
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Common Header | | Common Header |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Chunk #1 | | Chunk #1 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| ... | | ... |
skipping to change at page 13, line 36 skipping to change at page 91, line ?
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Adler-32 Checksum | | Adler-32 Checksum |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Source Port Number: 16 bit u_int Source Port Number: 16 bit u_int
This is the SCTP sender's port number. It can be used by the This is the SCTP sender's port number. It can be used by the
receiver, in combination with the source IP address, to identify the receiver, in combination with the source IP address, to identify the
association to which this datagram belongs. association to which this datagram belongs.
Desination Port Number: 16 bit u_int Destination Port Number: 16 bit u_int
This is the SCTP port number to which this datagram is destined. The This is the SCTP port number to which this datagram is destined. The
receiving endpoint will use this port number to de-multiplex the receiving host will use this port number to de-multiplex the
SCTP datagram to the correct receiving association/application. SCTP datagram to the correct receiving endpoint/application.
Stewart, et al [Page 12]
Verification Tag: 32 bit u_int Verification Tag: 32 bit u_int
The receiver of this datagram uses the Verification Tag to validate The receiver of this datagram uses the Verification Tag to validate
the sender of this SCTP datagram. On transmit, the value of this the sender of this SCTP datagram. On transmit, the value of this
Verification Tag MUST be set to the value of the Initiate Tag Verification Tag MUST be set to the value of the Initiate Tag
received from the peer endpoint during the association received from the peer endpoint during the association
initialization. initialization.
For datagrams carrying the INIT chunk, the transmitter MUST set the For datagrams carrying the INIT chunk, the transmitter MUST set the
skipping to change at page 15, line 9 skipping to change at page 91, line ?
Chunk ID: 8 bits, u_int Chunk ID: 8 bits, u_int
This field identifies the type of information contained in the Chunk This field identifies the type of information contained in the Chunk
Value field. It takes a value from 0x00 to 0xFF. The value of 0xFE Value field. It takes a value from 0x00 to 0xFF. The value of 0xFE
is reserved for vendor-specific extensions. The value of 0xFF is is reserved for vendor-specific extensions. The value of 0xFF is
reserved for future use as an extension field. Procedures for reserved for future use as an extension field. Procedures for
extending this field by vendors are defined in Section 2.4. extending this field by vendors are defined in Section 2.4.
The values of Chunk ID are defined as follows: The values of Chunk ID are defined as follows:
Stewart, et al [Page 13]
ID Value Chunk Type ID Value Chunk Type
----- ---------- ----- ----------
00000000 - Payload Data (DATA) 00000000 - Payload Data (DATA)
00000001 - Initiation (INIT) 00000001 - Initiation (INIT)
00000010 - Initiation Acknowledgment (INIT ACK) 00000010 - Initiation Acknowledgment (INIT ACK)
00000011 - Selective Acknowledgment (SACK) 00000011 - Selective Acknowledgment (SACK)
00000100 - Heartbeat Request (HEARTBEAT) 00000100 - Heartbeat Request (HEARTBEAT)
00000101 - Heartbeat Acknowledgment (HEARTBEAT ACK) 00000101 - Heartbeat Acknowledgment (HEARTBEAT ACK)
00000110 - Abort (ABORT) 00000110 - Abort (ABORT)
00000111 - Shutdown (SHUTDOWN) 00000111 - Shutdown (SHUTDOWN)
00001000 - Shutdown Acknowledgment (SHUTDOWN ACK) 00001000 - Shutdown Acknowledgment (SHUTDOWN ACK)
00001001 - Operation Error (ERROR) 00001001 - Operation Error (ERROR)
00001010 - Encryption Cookie (COOKIE) 00001010 - Encryption Cookie (COOKIE)
00001011 - Cookie Acknowledgement (COOKIE ACK) 00001011 - Cookie Acknowledgment (COOKIE ACK)
00001100 to 11111101 - reserved by IETF 00001100 to 11111101 - reserved by IETF
11111110 - Vendor-specific Chunk Extensions 11111110 - Vendor-specific Chunk Extensions
11111111 - IETF-defined Chunk Extensions 11111111 - IETF-defined Chunk Extensions
Chunk Flags: 8 bits Chunk Flags: 8 bits
The usage of these bits depends on the chunk type as given by the The usage of these bits depends on the chunk type as given by the
Chunk ID. Unless otherwise specified, they are set to zero on Chunk ID. Unless otherwise specified, they are set to zero on
transmit and are ignored on receipt. transmit and are ignored on receipt.
skipping to change at page 15, line 51 skipping to change at page 91, line ?
The Chunk Value field contains the actual information to be The Chunk Value field contains the actual information to be
transferred in the chunk. The usage and format of this field is transferred in the chunk. The usage and format of this field is
dependent on the Chunk ID. The Chunk Value field MUST be aligned on dependent on the Chunk ID. The Chunk Value field MUST be aligned on
32-bit boundaries. If the length of the chunk does not align on 32-bit boundaries. If the length of the chunk does not align on
32-bit boundaries, it is padded at the end with all zero octets. 32-bit boundaries, it is padded at the end with all zero octets.
SCTP defined chunks are described in detail in Section 2.3. The SCTP defined chunks are described in detail in Section 2.3. The
guideline for vendor-specific chunk extensions is discussed in Section guideline for vendor-specific chunk extensions is discussed in Section
2.4. And the guidelines for IETF-defined chunk extensions can be found 2.4. And the guidelines for IETF-defined chunk extensions can be found
in Section 11.1 of this document. in Section 12.1 of this document.
2.2.1 Optional/Variable-length Parameter Format 2.2.1 Optional/Variable-length Parameter Format
The optional and variable-length parameters contained in a chunk The optional and variable-length parameters contained in a chunk
are defined in a Type-Length-Value format as shown below. are defined in a Type-Length-Value format as shown below.
Stewart, et al [Page 14]
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Parameter Type | Parameter Length | | Parameter Type | Parameter Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
\ \ \ \
/ Parameter Value / / Parameter Value /
\ \ \ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
skipping to change at page 16, line 48 skipping to change at page 91, line ?
Parameter Value: variable-length. Parameter Value: variable-length.
The Value is dependent on the value of the Type field. The value The Value is dependent on the value of the Type field. The value
field MUST be aligned on 32-bit boundaries. If the value field is field MUST be aligned on 32-bit boundaries. If the value field is
not aligned on 32-bit boundaries it is padded at the end with all not aligned on 32-bit boundaries it is padded at the end with all
zero octets. The value field must be an integer number of octets. zero octets. The value field must be an integer number of octets.
The actual SCTP parameters are defined in the specific SCTP chunk The actual SCTP parameters are defined in the specific SCTP chunk
section. The guidelines for vendor-specific parameter extensions are section. The guidelines for vendor-specific parameter extensions are
discussed in Section 2.2.2. And the rules for IETF-defined parameter discussed in Section 2.2.2. And the rules for IETF-defined parameter
extensions are defined in Section 11.2. extensions are defined in Section 12.2.
2.2.2 Vendor-Specific Extension Parameter Format 2.2.2 Vendor-Specific Extension Parameter Format
This is to allow vendors to support their own extended parameters not This is to allow vendors to support their own extended parameters not
defined by the IETF. It MUST not affect the operation of SCTP. defined by the IETF. It MUST not affect the operation of SCTP.
Endpoints not equipped to interpret the vendor-specific information Endpoints not equipped to interpret the vendor-specific information
sent by a remote endpoint MUST ignore it (although it may be sent by a remote endpoint MUST ignore it (although it may be
reported). Endpoints that do not receive desired vendor-specific reported). Endpoints that do not receive desired vendor-specific
information SHOULD make an attempt to operate without it, although information SHOULD make an attempt to operate without it, although
they may do so (and report they are doing so) in a degraded mode. they may do so (and report they are doing so) in a degraded mode.
A summary of the Vendor-specific extension format is shown below. The A summary of the Vendor-specific extension format is shown below. The
fields are transmitted from left to right. fields are transmitted from left to right.
skipping to change at page 17, line 16 skipping to change at page 91, line ?
Endpoints not equipped to interpret the vendor-specific information Endpoints not equipped to interpret the vendor-specific information
sent by a remote endpoint MUST ignore it (although it may be sent by a remote endpoint MUST ignore it (although it may be
reported). Endpoints that do not receive desired vendor-specific reported). Endpoints that do not receive desired vendor-specific
information SHOULD make an attempt to operate without it, although information SHOULD make an attempt to operate without it, although
they may do so (and report they are doing so) in a degraded mode. they may do so (and report they are doing so) in a degraded mode.
A summary of the Vendor-specific extension format is shown below. The A summary of the Vendor-specific extension format is shown below. The
fields are transmitted from left to right. fields are transmitted from left to right.
Stewart, et al [Page 15]
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Parameter Type = 0xFFFE | Parameter Length | | Parameter Type = 0xFFFE | Parameter Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Vendor-Id | | Vendor-Id |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
\ \ \ \
/ Parameter Value / / Parameter Value /
\ \ \ \
skipping to change at page 18, line 21 skipping to change at page 91, line ?
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Parameter Type = 0xFFFE | Parameter Length | | Parameter Type = 0xFFFE | Parameter Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Vendor-Id | | Vendor-Id |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| VS-Type | VS-Length | | VS-Type | VS-Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
/ VS-Value / / VS-Value /
\ \ \ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Stewart, et al [Page 16]
VS-Type: 16 bit u_int VS-Type: 16 bit u_int
This field identifies the parameter included in the VS-Value field. This field identifies the parameter included in the VS-Value field.
It is assigned by the vendor. It is assigned by the vendor.
VS-Length: 16 bit u_int VS-Length: 16 bit u_int
This field is the length of the vendor-specific parameter and This field is the length of the vendor-specific parameter and
Includes the VS-Type, VS-Length and VS-Value (if included) fields. Includes the VS-Type, VS-Length and VS-Value (if included) fields.
skipping to change at page 19, line 34 skipping to change at page 91, line ?
noted, each parameter MUST only be included once in the INIT chunk. noted, each parameter MUST only be included once in the INIT chunk.
Fixed Parameters Status Fixed Parameters Status
---------------------------------------------- ----------------------------------------------
Initiate Tag Mandatory Initiate Tag Mandatory
Receiver Window Credit Mandatory Receiver Window Credit Mandatory
Number of Outbound Streams Mandatory Number of Outbound Streams Mandatory
Number of Inbound Streams Mandatory Number of Inbound Streams Mandatory
Initial TSN Mandatory Initial TSN Mandatory
Stewart, et al [Page 17]
Variable Parameters Status Type Value Variable Parameters Status Type Value
------------------------------------------------------------- -------------------------------------------------------------
IPv4 Address (Note 1) Optional 0x0005 IPv4 Address (Note 1) Optional 0x0005
IPv6 Address (Note 1) Optional 0x0006 IPv6 Address (Note 1) Optional 0x0006
Cookie Preservative Optional 0x0009 Cookie Preservative Optional 0x0009
Note 1: The INIT chunks may contain multiple addresses that may be Note 1: The INIT chunks may contain multiple addresses that may be
IPv4 and/or IPv6 in any combination. IPv4 and/or IPv6 in any combination.
Chunk Flags field in INIT is reserved, and all bits in it should be Chunk Flags field in INIT is reserved, and all bits in it should be
skipping to change at page 20, line 41 skipping to change at page 91, line ?
Defines the maximum number of streams the sender of this INIT chunk Defines the maximum number of streams the sender of this INIT chunk
allows the peer end to create in this association. The value 0 MUST allows the peer end to create in this association. The value 0 MUST
NOT be used. NOT be used.
Initial TSN (I-TSN) : 32 bit u_int Initial TSN (I-TSN) : 32 bit u_int
Defines the initial TSN that the sender will use. The valid range is Defines the initial TSN that the sender will use. The valid range is
from 0x0 to 0xffffffff. This field MAY be set to the value of the from 0x0 to 0xffffffff. This field MAY be set to the value of the
Initiate Tag field. Initiate Tag field.
Stewart, et al [Page 18]
Vendor-specific parameters are allowed in INIT. However, they MUST be Vendor-specific parameters are allowed in INIT. However, they MUST be
appended to the end of the above INIT chunks. The format of the appended to the end of the above INIT chunks. The format of the
vendor-specific parameters MUST follow the Type-Length-value format as vendor-specific parameters MUST follow the Type-Length-value format as
defined in Section 2.2.2. In case an endpoint does not support the defined in Section 2.2.2. In case an endpoint does not support the
vendor-specific chunks received, it MUST ignore them. vendor-specific chunks received, it MUST ignore them.
2.3.1.1 Optional/Variable Length Parameters in INIT 2.3.1.1 Optional/Variable Length Parameters in INIT
The following parameters follow the Type-Length-Value format as The following parameters follow the Type-Length-Value format as
defined in Section 2.2.1. The IP address fields MUST come after defined in Section 2.2.1. The IP address fields MUST come after
skipping to change at page 22, line 10 skipping to change at page 91, line ?
IPv6 Address: 128 bit IPv6 Address: 128 bit
Contains an IPv6 address of the sending endpoint. It is binary Contains an IPv6 address of the sending endpoint. It is binary
encoded. encoded.
Combining with the Source Port Number in the SCTP common header, the Combining with the Source Port Number in the SCTP common header, the
value passed in an IPv4 or IPv6 Address parameter indicates a value passed in an IPv4 or IPv6 Address parameter indicates a
transport address the sender of the INIT will support for the transport address the sender of the INIT will support for the
association being initiated. That is, during the lifetime of this association being initiated. That is, during the lifetime of this
association, this IP address may appear in the source address field association, this IP address may appear in the source address field
Stewart, et al [Page 19]
of a datagram sent from the sender of the INIT, and may be used as a of a datagram sent from the sender of the INIT, and may be used as a
destination address of a datagram sent from the receiver of the destination address of a datagram sent from the receiver of the
INIT. INIT.
More than one IP Address parameters can be included in an INIT More than one IP Address parameters can be included in an INIT
chunk when the INIT sender is multi-homed. Moreover, a multi-homed chunk when the INIT sender is multi-homed. Moreover, a multi-homed
endpoint may have access to different types of network, thus more endpoint may have access to different types of network, thus more
than one address type may be present in one INIT chunk, i.e., IPv4 than one address type may be present in one INIT chunk, i.e., IPv4
and IPv6 transport addresses are allowed in the same INIT message. and IPv6 transport addresses are allowed in the same INIT message.
If the INIT contains a least one IP Address parameter, then only the If the INIT contains at least one IP Address parameter, then only the
transport address(es) provided within the INIT may be used as transport address(es) provided within the INIT may be used as
destinations by the responding end. If the INIT does not contain any destinations by the responding end. If the INIT does not contain any
IP Address parameters, the responding end MUST use the source IP Address parameters, the responding end MUST use the source
address associated with the received SCTP datagram as its sole address associated with the received SCTP datagram as its sole
destination address for the association. destination address for the association.
Cookie Preservative Cookie Preservative
The sender of the INIT shall use this parameter to suggest to the The sender of the INIT shall use this parameter to suggest to the
receiver of the INIT for a longer life-span of the Encryption Cookie. receiver of the INIT for a longer life-span of the Encryption Cookie.
skipping to change at page 22, line 43 skipping to change at page 91, line ?
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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1|0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0| |0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1|0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Suggested Cookie Life-span Increment (msec.) | | Suggested Cookie Life-span Increment (msec.) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Suggested Cookie Life-span Increment: 32bit u_int Suggested Cookie Life-span Increment: 32bit u_int
This parameter indicates to the receiver how much increment the This parameter indicates to the receiver how much increment the
sender wishes the reciever to add to its default cookie life-span. sender wishes the receiver to add to its default cookie life-span.
This optional parameter should be added to the INIT message by the This optional parameter should be added to the INIT message by the
sender when it re-attempts establishing an association with a peer sender when it re-attempts establishing an association with a peer
to which its first attempt of establishing the association failed to which its previous attempt of establishing the association failed
due to a Stale COOKIE error. Note, the receiver MAY choose to ignore due to a Stale COOKIE error. Note, the receiver MAY choose to ignore
the suggested cookie life-span increase for its own security the suggested cookie life-span increase for its own security
reasons. reasons.
2.3.2 Initiation Acknowledgement (INIT ACK) (00000010): 2.3.2 Initiation Acknowledgment (INIT ACK) (00000010):
The INIT ACK chunk is used to acknowledge the initiation of an SCTP The INIT ACK chunk is used to acknowledge the initiation of an SCTP
association. association.
The parameter part of INIT ACK is formatted similarly to the INIT The parameter part of INIT ACK is formatted similarly to the INIT
chunk. It uses two extra variable parameters: The Encryption Cookie chunk. It uses two extra variable parameters: The Encryption Cookie
and the Unrecognized Parameter: and the Unrecognized Parameter:
The format of the INIT ACK message is shown below: The format of the INIT ACK message is shown below:
Stewart, et al [Page 20]
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0 0 0 0 0 0 1 0| Chunk Flags | Chunk Length | |0 0 0 0 0 0 1 0| Chunk Flags | Chunk Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Initiate Tag | | Initiate Tag |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Receiver Window Credit | | Receiver Window Credit |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Number of Outbound Streams | Number of Inbound Streams | | Number of Outbound Streams | Number of Inbound Streams |
skipping to change at page 24, line 28 skipping to change at page 91, line ?
the sender of the INIT ACK for the lifetime of the association being the sender of the INIT ACK for the lifetime of the association being
initiated. initiated.
If the INIT ACK contains at least one IP Address parameter, then only If the INIT ACK contains at least one IP Address parameter, then only
the transport address(es) explicitly indicated in the INIT ACK may be the transport address(es) explicitly indicated in the INIT ACK may be
used as the destination(s) by the receiver of the INIT ACK. However, used as the destination(s) by the receiver of the INIT ACK. However,
if the INIT ACK contains no IP Address parameter, the receiver of the if the INIT ACK contains no IP Address parameter, the receiver of the
INIT ACK MUST take the source IP address associated with this INIT ACK INIT ACK MUST take the source IP address associated with this INIT ACK
as its sole destination address for this association. as its sole destination address for this association.
Stewart, et al [Page 21]
The Encryption Cookie and Unrecognized Parameters use the Type-Length- The Encryption Cookie and Unrecognized Parameters use the Type-Length-
Value format as defined in Section 2.2.1 and are described below. The Value format as defined in Section 2.2.1 and are described below. The
other fields are defined the same as their counterparts in the INIT other fields are defined the same as their counterparts in the INIT
message. message.
2.3.2.1 Optional or Variable Length Parameters 2.3.2.1 Optional or Variable Length Parameters
Encryption Cookie: variable size, depending on Size of Cookie Encryption Cookie: variable size, depending on Size of Cookie
This field MUST contain all the necessary state and parameter This field MUST contain all the necessary state and parameter
skipping to change at page 24, line 54 skipping to change at page 91, line ?
Unrecognized Parameters: Variable Size. Unrecognized Parameters: Variable Size.
This parameter is returned to the originator of the INIT message if This parameter is returned to the originator of the INIT message if
the receiver does not recognize one or more Optional TLV parameters the receiver does not recognize one or more Optional TLV parameters
in the INIT chunk. This parameter field will contain the in the INIT chunk. This parameter field will contain the
unrecognized parameters copied from the INIT message complete unrecognized parameters copied from the INIT message complete
with TLV. with TLV.
Vendor-Specific parameters are allowed in INIT ACK. However, they Vendor-Specific parameters are allowed in INIT ACK. However, they
MUST be defined using the format described in Section 2.2.2, and be MUST be defined using the format described in Section 2.2.2, and be
appended to the end of the above INIT ACK chunks. In case the receiver appended to the end of the above INIT ACK chunk. In case the receiver
of the INIT ACK does not support the vendor-specific parameters of the INIT ACK does not support the vendor-specific parameters
received, it MUST ignore those fields. received, it MUST ignore those fields.
2.3.3 Selective Acknowledgement (SACK) (00000011): 2.3.3 Selective Acknowledgment (SACK) (00000011):
This chunk is sent to the remote endpoint to acknowledge received DATA This chunk is sent to the remote endpoint to acknowledge received DATA
chunks and to inform the remote endpoint of gaps in the received chunks and to inform the remote endpoint of gaps in the received
subsequences of DATA chunks as represented by their TSNs. subsequences of DATA chunks as represented by their TSNs.
The SACK MUST contain the Cumulative TSN ACK and Advertised Receiver The SACK MUST contain the Cumulative TSN ACK and Advertised Receiver
Window Credit (a_rwnd) parameters. By definition, the value of the Window Credit (a_rwnd) parameters. By definition, the value of the
Cumulative TSN ACK parameter is the last TSN received at the time the Cumulative TSN ACK parameter is the last TSN received at the time the
Selective ACK is sent, before a break in the sequence of received TSNs Selective ACK is sent, before a break in the sequence of received TSNs
occurs; the next TSN value following this one has not yet been occurs; the next TSN value following this one has not yet been
received at the reporting end. This parameter therefore acknowledges received at the reporting end. This parameter therefore acknowledges
receipt of all TSNs up to and including the value given. receipt of all TSNs up to and including the value given.
The handling of the a_rwnd by the receiver of the SACK is diccussed in The handling of the a_rwnd by the receiver of the SACK is discussed in
detail in Section 5.2.1. detail in Section 5.2.1.
The Selective ACK also contains zero or more fragment reports. Each The Selective ACK also contains zero or more fragment reports. Each
fragment report acknowledges a subsequence of TSNs received following fragment report acknowledges a subsequence of TSNs received following
a break in the sequence of received TSNs. By definition, all TSNs a break in the sequence of received TSNs. By definition, all TSNs
acknowledged by fragment reports are higher than the value of the acknowledged by fragment reports are higher than the value of the
Cumulative TSN ACK. Cumulative TSN ACK.
Stewart, et al [Page 22]
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0 0 0 0 0 0 1 1|Chunk Flags | Chunk Length | |0 0 0 0 0 0 1 1|Chunk Flags | Chunk Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cumulative TSN ACK | | Cumulative TSN ACK |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Advertised Receiver Window Credit (a_rwnd) | | Advertised Receiver Window Credit (a_rwnd) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Number of Fragments = N | (Reserved) | | Number of Fragments = N | (Reserved) |
skipping to change at page 26, line 4 skipping to change at page 91, line ?
Chunk Flags: Chunk Flags:
Set to all zeros on transmit and ignored on receipt. Set to all zeros on transmit and ignored on receipt.
Cumulative TSN ACK: 32 bit u_int Cumulative TSN ACK: 32 bit u_int
This parameter contains the TSN of the last DATA chunk received in This parameter contains the TSN of the last DATA chunk received in
sequence before a gap. sequence before a gap.
Advertised Receiver Window Credit (a_rwnd): 32 bit u_int Advertised Receiver Window Credit (a_rwnd): 32 bit u_int
This field indicates the updated receive buffer space in octets of This field indicates the updated receive buffer space in octets of
the sender of this SACK, see Section 5.11 for details. the sender of this SACK, see Section 5.2.1 for details.
Number of Fragments: 16 bit u_int Number of Fragments: 16 bit u_int
Indicates the number of TSN fragments included in this Selective Indicates the number of TSN fragments included in this Selective
ACK. ACK.
Reserved: 16 bit Reserved: 16 bit
Must be set to all 0 by the sender and ignored by the receiver. Must be set to all 0 by the sender and ignored by the receiver.
Fragments: Fragments:
These fields contain the ack fragments. They are repeated for each These fields contain the ack fragments. They are repeated for each
fragment up to the number of fragments defined in the Number of fragment up to the number of fragments defined in the Number of
Fragments field. All DATA chunks with TSNs between the (Cumulative Fragments field. All DATA chunks with TSNs between the (Cumulative
TSN ACK + Fragment Start) and (Cumulative TSN ACK + Fragment End) of TSN ACK + Fragment Start) and (Cumulative TSN ACK + Fragment End) of
each fragment are assumed to have been received correctly. each fragment are assumed to have been received correctly.
Stewart, et al [Page 23]
Fragment Start: 16 bit u_int Fragment Start: 16 bit u_int
Indicates the Start offset TSN for this fragment. To calculate the Indicates the Start offset TSN for this fragment. To calculate the
actual TSN number the Cumulative TSN ACK is added to this actual TSN number the Cumulative TSN ACK is added to this
offset number to yield the TSN. This calculated TSN identifies offset number to yield the TSN. This calculated TSN identifies
the first TSN in this fragment that has been received. the first TSN in this fragment that has been received.
Fragment End: 16 bit u_int Fragment End: 16 bit u_int
Indicates the End offset TSN for this fragment. To calculate the Indicates the End offset TSN for this fragment. To calculate the
skipping to change at page 27, line 37 skipping to change at page 91, line ?
----------------+--------------- ----------------+---------------
| a_rwnd = 0x1234 | | a_rwnd = 0x1234 |
----------------+--------------- ----------------+---------------
| num of frag=2 | (rev = 0) | | num of frag=2 | (rev = 0) |
----------------+--------------- ----------------+---------------
|frag #1 strt=2 |frag #1 end=3 | |frag #1 strt=2 |frag #1 end=3 |
----------------+--------------- ----------------+---------------
|frag #2 strt=5 |frag #2 end=5 | |frag #2 strt=5 |frag #2 end=5 |
-------------------------------- --------------------------------
Stewart, et al [Page 24]
2.3.4 Heartbeat Request (HEARTBEAT) (00000100): 2.3.4 Heartbeat Request (HEARTBEAT) (00000100):
An endpoint should send this chunk to its peer endpoint of the current An endpoint should send this chunk to its peer endpoint of the current
association to probe the reachability of a particular destination association to probe the reachability of a particular destination
transport address defined in the present association. transport address defined in the present association.
The parameter field contains the Heartbeat Information which is a The parameter field contains the Heartbeat Information which is a
variable length opeque data structure understood only by the sender. variable length opaque data structure understood only by the sender.
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0 0 0 0 0 1 0 0| Chunk Flags | Heartbeat Length | |0 0 0 0 0 1 0 0| Chunk Flags | Heartbeat Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
\ \ \ \
/ Heartbeat Information (Variable-Length) / / Heartbeat Information (Variable-Length) /
\ \ \ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
skipping to change at page 28, line 43 skipping to change at page 91, line ?
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
/ Sender-specific Heartbeat Info / / Sender-specific Heartbeat Info /
\ \ \ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
The Sender-specific Heartbeat Info field should normally include The Sender-specific Heartbeat Info field should normally include
information about the sender's current time when this HEARTBEAT information about the sender's current time when this HEARTBEAT
message is sent and the destination transport address to which this message is sent and the destination transport address to which this
HEARTBEAT is sent (see Section 7.3). HEARTBEAT is sent (see Section 7.3).
Stewart, et al [Page 25]
2.3.5 Heartbeat Acknowledgment (HEARTBEAT ACK) (00000101): 2.3.5 Heartbeat Acknowledgment (HEARTBEAT ACK) (00000101):
An endpoint should send this chunk to its peer endpoint as a response An endpoint should send this chunk to its peer endpoint as a response
to a Heartbeat Request (see Section 7.3). to a Heartbeat Request (see Section 7.3).
The parameter field contains a variable length opeque data structure. The parameter field contains a variable length opaque data structure.
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0 0 0 0 0 1 0 1| Chunk Flags | Heartbeat Ack Length | |0 0 0 0 0 1 0 1| Chunk Flags | Heartbeat Ack Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
\ \ \ \
/ Heartbeat Information (Variable-Length) / / Heartbeat Information (Variable-Length) /
\ \ \ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
skipping to change at page 29, line 28 skipping to change at page 91, line ?
Heartbeat Ack Length: Heartbeat Ack Length:
Set to the size of the chunk in octets, including the chunk header Set to the size of the chunk in octets, including the chunk header
and the Heartbeat Information field. and the Heartbeat Information field.
Heartbeat Information: Heartbeat Information:
The values of this field SHALL be copied from the Heartbeat The values of this field SHALL be copied from the Heartbeat
Information field found in the Heartbeat Request to which this Information field found in the Heartbeat Request to which this
Heartbeat Acknowledgement is responding. Heartbeat Acknowledgment is responding.
2.3.6 Abort Association (ABORT) (00000110): 2.3.6 Abort Association (ABORT) (00000110):
The ABORT chunk is sent to the peer of an association to terminate the The ABORT chunk is sent to the peer of an association to terminate the
association. The ABORT chunk has no parameters. The sender of ABORT association. The ABORT chunk may contain cause parameters to inform
may bundle one or more ERROR chunks to indicate the reason of abort. the receiver the reason of the abort. DATA chunks MUST not be bundled
However, if ERROR chunks are bundled with an ABORT, they MUST be with ABORT. Control chunks MAY be bundled with an ABORT but they MUST
placed before the ABORT chunk in the outbound datagram. be placed before the ABORT in the SCTP datagram, or they will be
ignored by the receiver.
If an endpoint receives an ABORT with a format error or for an If an endpoint receives an ABORT with a format error or for an
association that doesn't exist, it MUST silently discard it. association that doesn't exist, it MUST silently discard it.
Moreover, under any circumstances, an endpoint that receives an ABORT
MUST never respond to that ABORT by sending an ABORT of its own.
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0 0 0 0 0 1 1 0|Chunk Flags |0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0| |0 0 0 0 0 1 1 0| Chunk Flags | Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
\ \
/ zero or more Error Causes /
\ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Stewart, et al [Page 26]
Chunk Flags: Chunk Flags:
Set to zero on transmit and ignored on receipt. Set to zero on transmit and ignored on receipt.
Some special rules apply to the Verification Tag field of SCTP Length:
Set to the size of the chunk in octets, including the chunk header
and all the Error Cause fields present.
See Section 2.3.9 for Error Cause definitions.
Note: Special rules apply to the Verification Tag field of SCTP
datagrams which carry an ABORT, see Section 7.5 for details. datagrams which carry an ABORT, see Section 7.5 for details.
2.3.7 SHUTDOWN (00000111): 2.3.7 SHUTDOWN (00000111):
An endpoint in an association MUST use this chunk to initiate a An endpoint in an association MUST use this chunk to initiate a
graceful termination of the association with its peer. This chunk has graceful termination of the association with its peer. This chunk has
the following format. the following format.
0 1 2 3 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 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
skipping to change at page 30, line 28 skipping to change at page 91, line ?
Chunk Flags: Chunk Flags:
Set to zero on transmit and ignored on receipt. Set to zero on transmit and ignored on receipt.
Cumulative TSN ACK: 32 bit u_int Cumulative TSN ACK: 32 bit u_int
This parameter contains the TSN of the last chunk received in This parameter contains the TSN of the last chunk received in
sequence before any gaps. sequence before any gaps.
Stewart, et al [Page 27]
2.3.8 Shutdown Acknowledgment (SHUTDOWN ACK) (00001000): 2.3.8 Shutdown Acknowledgment (SHUTDOWN ACK) (00001000):
This chunk MUST be used to acknowledge the receipt of the SHUTDOWN This chunk MUST be used to acknowledge the receipt of the SHUTDOWN
chunk at the completion of the shutdown process, see Section 8.2 for chunk at the completion of the shutdown process, see Section 8.2 for
details. details.
The SHUTDOWN ACK chunk has no parameters. The SHUTDOWN ACK chunk has no parameters.
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0 0 0 0 1 0 0 0|Chunk Flags |0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0| |0 0 0 0 1 0 0 0|Chunk Flags |0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Chunk Flags: Chunk Flags:
Set to zero on transmit and ignored on receipt. Set to zero on transmit and ignored on receipt.
Note: if the endpoint that receives the SHUTDOWN message does not have Note: if the endpoint that receives the SHUTDOWN message does not have
a TCB or tag for the sender of the SHUTDOWN, the receiver SHALL still a TCB or tag for the sender of the SHUTDOWN, the receiver MUST still
respond. In such cases, the receiver SHALL send back a stand-alone respond. In such cases, the receiver MUST send back a stand-alone
SHUTDOWN ACK chunk in an SCTP datagram with the Verification Tag field SHUTDOWN ACK chunk in an SCTP datagram with the Verification Tag field
of the common header filled with all '0's. of the common header filled with all '0's.
2.3.9 Operation Error (ERROR) (00001001): 2.3.9 Operation Error (ERROR) (00001001):
This chunk is sent to the other endpoint in the association to notify This chunk is sent to the other endpoint in the association to notify
certain error conditions. It contains one or more error causes. It has certain error conditions. It contains one or more error causes. It has
the following parameters: the following parameters:
0 1 2 3 0 1 2 3
skipping to change at page 31, line 33 skipping to change at page 91, line ?
Set to zero on transmit and ignored on receipt. Set to zero on transmit and ignored on receipt.
Length: Length:
Set to the size of the chunk in octets, including the chunk header Set to the size of the chunk in octets, including the chunk header
and all the Error Cause fields present. and all the Error Cause fields present.
Error causes are defined as variable-length parameters using the Error causes are defined as variable-length parameters using the
format described in 2.2.1, i.e.: format described in 2.2.1, i.e.:
Stewart, et al [Page 28]
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cause Code | Cause Length | | Cause Code | Cause Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
/ Cause-specific Information / / Cause-specific Information /
\ \ \ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Cause Code: 16 bit u_int Cause Code: 16 bit u_int
skipping to change at page 32, line 35 skipping to change at page 91, line ?
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Number of missing params=N | | Number of missing params=N |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Missing Param Type #1 | Missing Param Type #2 | | Missing Param Type #1 | Missing Param Type #2 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Missing Param Type #N-1 | Missing Param Type #N | | Missing Param Type #N-1 | Missing Param Type #N |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Each missing mandatory parameter type should be specified. Each missing mandatory parameter type should be specified.
Stewart, et al [Page 29]
Cause of error Cause of error
-------------- --------------
Stale Cookie Error: indicating the receiving of a valid cookie Stale Cookie Error: indicating the receiving of a valid cookie
which is however expired. which is however expired.
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cause Code=3 | Cause Length=8 | | Cause Code=3 | Cause Length=8 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Measure of Staleness (usec.) | | Measure of Staleness (usec.) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
The sender of this error cause MAY choose to report how long past The sender of this error cause MAY choose to report how long past
expiration the cookie is, by putting in the Measure of Staleness expiration the cookie is, by putting in the Measure of Staleness
field the difference, in microseconds, between the current time and field the difference, in microseconds, between the current time and
the time the cookie expired. If the sender does not wish to provide the time the cookie expired. If the sender does not wish to provide
this information it should set Measure of staleness to 0. this information it should set Measure of staleness to 0.
Cause of error Cause of error
--------------- ---------------
Out of Resource: indicating that the sender is out of resource. This Out of Resource: indicating that the sender is out of resource. This
is usually sent in combination with an ABORT. is usually sent in combination with or within an ABORT.
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cause Code=4 | Cause Length=4 | | Cause Code=4 | Cause Length=4 |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Guidelines for IETF-defined Error Cause extensions are discussed in Guidelines for IETF-defined Error Cause extensions are discussed in
Section 11.3 of this document. Section 12.3 of this document.
2.3.10 Encryption Cookie (COOKIE) (00001010): 2.3.10 Encryption Cookie (COOKIE) (00001010):
This chunk is used only during the initialization of an association. This chunk is used only during the initialization of an association.
It is sent by the initiator of an association to its peer to complete It is sent by the initiator of an association to its peer to complete
the initialization process. This chunk MUST precede any DATA chunk the initialization process. This chunk MUST precede any chunk
sent within the association, but MAY be bundled with one or more DATA sent within the association, but MAY be bundled with one or more DATA
chunks in the same datagram. chunks in the same datagram.
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0 0 0 0 1 0 1 0|Chunk Flags | Length | |0 0 0 0 1 0 1 0|Chunk Flags | Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Cookie | | Cookie |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Chunk Flags: 8 bit Chunk Flags: 8 bit
Set to zero on transmit and ignored on receipt. Set to zero on transmit and ignored on receipt.
Length: 16 bit u_int Length: 16 bit u_int
Set to the size of the chunk in octets, including the 4 octets of Set to the size of the chunk in octets, including the 4 octets of
the chunk header and the size of the Cookie. the chunk header and the size of the Cookie.
Stewart, et al [Page 30]
Cookie: variable size Cookie: variable size
This field must contain the exact cookie received in a previous INIT This field must contain the exact cookie received in a previous INIT
ACK. ACK.
2.3.11 Cookie Acknowledgment (COOKIE ACK) (00001011): 2.3.11 Cookie Acknowledgment (COOKIE ACK) (00001011):
This chunk is used only during the initialization of an association. This chunk is used only during the initialization of an association.
It is used to acknowledge the receipt of a COOKIE chunk. This chunk It is used to acknowledge the receipt of a COOKIE chunk. This chunk
MUST precede any DATA chunk sent within the association, but MAY be MUST precede any chunk sent within the association, but MAY be
bundled with one or more DATA chunks in the same SCTP datagram. bundled with one or more DATA chunks in the same SCTP datagram.
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|0 0 0 0 1 0 1 1|Chunk Flags |0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0| |0 0 0 0 1 0 1 1|Chunk Flags |0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0|
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Chunk Flags: Chunk Flags:
skipping to change at page 34, line 34 skipping to change at page 91, line ?
Reserved: 5 bits Reserved: 5 bits
should be set to all '0's and ignored by the receiver. should be set to all '0's and ignored by the receiver.
U bit: 1 bit U bit: 1 bit
The (U)nordered bit, if set, indicates that this is an unordered The (U)nordered bit, if set, indicates that this is an unordered
data chunk, and there is NO Stream Sequence Number assigned to this data chunk, and there is NO Stream Sequence Number assigned to this
DATA chunk. Therefore, the receiver MUST ignore the Stream Sequence DATA chunk. Therefore, the receiver MUST ignore the Stream Sequence
Number field. Number field.
After reassembly (if necessary), unordered data chunks MUST be Stewart, et al [Page 31]
After re-assembly (if necessary), unordered data chunks MUST be
dispatched to the upper layer by the receiver without any attempt of dispatched to the upper layer by the receiver without any attempt of
re-ordering. re-ordering.
Note, if an unordered user message is segmented, each segment of the Note, if an unordered user message is segmented, each segment of the
message MUST have its U bit set to 1. message MUST have its U bit set to 1.
B bit: 1 bit B bit: 1 bit
The (B)eginning segment bit, if set, indicates the first segment of The (B)eginning segment bit, if set, indicates the first segment of
a user message. a user message.
skipping to change at page 35, line 32 skipping to change at page 91, line ?
TSN : 32 bits (32 bit u_int) TSN : 32 bits (32 bit u_int)
This value represents the TSN for this DATA chunk. The valid range This value represents the TSN for this DATA chunk. The valid range
of TSN is from 0x0 to 0xffffffff. of TSN is from 0x0 to 0xffffffff.
Stream Identifier S: 16 bit u_int Stream Identifier S: 16 bit u_int
Identifies the stream to which the following user data belongs. Identifies the stream to which the following user data belongs.
Sequence Number n: 16 bit u_int Stream Sequence Number n: 16 bit u_int
This value presents the sequence number of the following user This value presents the stream sequence number of the following user
data within the stream S. Valid range is 0x0 to 0xFFFF. data within the stream S. Valid range is 0x0 to 0xFFFF.
Note, when a user message is segmented by SCTP for transport, the Note, when a user message is segmented by SCTP for transport, the
same sequence number MUST be carried in each of the segments of same stream sequence number MUST be carried in each of the segments of
the message. the message.
Stewart, et al [Page 32]
Payload Protocol Identifier: 32 bits (32 bit u_int) Payload Protocol Identifier: 32 bits (32 bit u_int)
This value represents an application (or upper layer) specified This value represents an application (or upper layer) specified
protocol identifier. This value is passed to SCTP by its upper layer protocol identifier. This value is passed to SCTP by its upper layer
and sent to its peer. This identifier is not used by SCTP but may be and sent to its peer. This identifier is not used by SCTP but may be
used by certain network entities as well as the peer application to used by certain network entities as well as the peer application to
identify the type of information being carried in this DATA chunk. identify the type of information being carried in this DATA chunk.
The value 0x0 indicates no application identifier is specified by The value 0x0 indicates no application identifier is specified by
the upper layer for this payload data. the upper layer for this payload data.
User Data: variable length User Data: variable length
This is the payload user data. The implementation MUST pad the end This is the payload user data. The implementation MUST pad the end
of the data to a 32 bit boundary with 0 octets. Any padding should of the data to a 32 bit boundary with 0 octets. Any padding MUST
NOT be included in the length field. NOT be included in the length field.
2.4 Vendor-Specific Chunk Extensions 2.4 Vendor-Specific Chunk Extensions
This Chunk type is available to allow vendors to support their own This Chunk type is available to allow vendors to support their own
extended data formats not defined by the IETF. It MUST not affect the extended data formats not defined by the IETF. It MUST not affect the
operation of SCTP. In particular, when adding a Vendor Specific chunk operation of SCTP. In particular, when adding a Vendor Specific chunk
type, the vendor defined chunks MUST obey the congestion avoidance type, the vendor defined chunks MUST obey the congestion avoidance
rules defined in this document if they carry user data. User data is rules defined in this document if they carry user data. User data is
defined as any data transported over the association that is delivered defined as any data transported over the association that is delivered
skipping to change at page 36, line 29 skipping to change at page 91, line ?
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
\ \ \ \
/ Value / / Value /
\ \ \ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Type: 8 bit u_int Type: 8 bit u_int
0xFE for all Vendor-Specific chunks. 0xFE for all Vendor-Specific chunks.
Stewart, et al [Page 33]
Flags: 8 bit u_int Flags: 8 bit u_int
Vendor specific flags. Vendor specific flags.
Length: 16 bit u_int Length: 16 bit u_int
Size of this Vendor-Specific chunks in octets, including the Type, Size of this Vendor-Specific chunks in octets, including the Type,
Flags, Length, Vendor-Id, and Value fields. Flags, Length, Vendor-Id, and Value fields.
Vendor-Id: 32 bit u_int Vendor-Id: 32 bit u_int
skipping to change at page 37, line 27 skipping to change at page 91, line ?
The state diagram in the figures below illustrates state changes, The state diagram in the figures below illustrates state changes,
together with the causing events and resulting actions. Note that some together with the causing events and resulting actions. Note that some
of the error conditions are not shown in the state diagram. Full of the error conditions are not shown in the state diagram. Full
description of all special cases should be found in the text. description of all special cases should be found in the text.
Note, chunk names are given in all capital letters, while parameter Note, chunk names are given in all capital letters, while parameter
names have the first letter capitalized, e.g., COOKIE chunk type vs. names have the first letter capitalized, e.g., COOKIE chunk type vs.
Cookie parameter. Cookie parameter.
Stewart, et al [Page 34]
----- -------- (frm any state) ----- -------- (frm any state)
/ \ / rcv ABORT [ABORT] / \ / rcv ABORT [ABORT]
rcv INIT | | | ---------- or ---------- rcv INIT | | | ---------- or ----------
--------------- | v v delete TCB snd ABORT --------------- | v v delete TCB snd ABORT
generate Cookie \ +---------+ delete TCB generate Cookie \ +---------+ delete TCB
snd INIT.ACK ---| CLOSED | snd INIT.ACK ---| CLOSED |
+---------+ +---------+
/ \ [ASSOCIATE] / \ [ASSOCIATE]
/ \ --------------- / \ ---------------
| | create TCB | | create TCB
skipping to change at page 38, line 31 skipping to change at page 91, line ?
+---------+ | +---------+ |
|SHUTDOWN | | rcv SHUTDOWN |SHUTDOWN | | rcv SHUTDOWN
|PENDING | | ---------------- |PENDING | | ----------------
+---------+ | x +---------+ | x
| | | |
No more outstanding | | No more outstanding | |
------------------- | | ------------------- | |
snd SHUTDOWN | | snd SHUTDOWN | |
strt shutdown timer | | strt shutdown timer | |
v v v v
tewart, et al [Page 35]
+---------+ +-----------+ +---------+ +-----------+
(4) |SHUTDOWN | | SHUTDOWN | (5) (4) |SHUTDOWN | | SHUTDOWN | (5)
|SENT | | RECEIVED | |SENT | | RECEIVED |
+---------+ +-----------+ +---------+ +-----------+
| | | |
rcv SHUTDOWN.ACK | | x rcv SHUTDOWN.ACK | | x
------------------- | |----------------- ------------------- | |-----------------
stop shutdown timer | | retransmit missing DATA stop shutdown timer | | retransmit missing DATA
delete TCB | | send SHUTDOWN.ACK delete TCB | | send SHUTDOWN.ACK
| | delete TCB | | delete TCB
skipping to change at page 39, line 30 skipping to change at page 91, line ?
(5) In SHUTDOWN-RECEIVED state, the endpoint MUST NOT accept any new (5) In SHUTDOWN-RECEIVED state, the endpoint MUST NOT accept any new
send request from its SCTP user. send request from its SCTP user.
4. Association Initialization 4. Association Initialization
Before the first data transmission can take place from one SCTP Before the first data transmission can take place from one SCTP
endpoint ("A") to another SCTP endpoint ("Z"), the two endpoints must endpoint ("A") to another SCTP endpoint ("Z"), the two endpoints must
complete an initialization process in order to set up an SCTP complete an initialization process in order to set up an SCTP
association between them. association between them.
The SCTP user at an endpoint SHOULD use the ASSOCIATE primitive to The SCTP user at an endpoint should use the ASSOCIATE primitive to
initialize an SCTP association to another SCTP endpoint. initialize an SCTP association to another SCTP endpoint.
Stewart, et al [Page 36]
IMPLEMENTATION NOTE: From an SCTP-user's point of view, an IMPLEMENTATION NOTE: From an SCTP-user's point of view, an
association may be implicitly opened, without an ASSOCIATE primitive association may be implicitly opened, without an ASSOCIATE primitive
(see 9.1 B) being invoked, by the initiating endpoint's sending of (see 9.1 B) being invoked, by the initiating endpoint's sending of
the first user data to the destination endpoint. The initiating SCTP the first user data to the destination endpoint. The initiating SCTP
will assume default values for all mandatory and optional parameters will assume default values for all mandatory and optional parameters
for the INIT/INIT ACK. for the INIT/INIT ACK.
Once the association is established, unidirectional streams will be Once the association is established, unidirectional streams will be
open for data transfer on both ends (see Section 4.1.1). open for data transfer on both ends (see Section 4.1.1).
skipping to change at page 40, line 32 skipping to change at page 91, line ?
received in the INIT ACK message in a cookie chunk, restart the received in the INIT ACK message in a cookie chunk, restart the
T1-init timer, and enter the COOKIE-SENT state. T1-init timer, and enter the COOKIE-SENT state.
Note, the cookie chunk can be bundled with any pending outbound Note, the cookie chunk can be bundled with any pending outbound
DATA chunks, but it MUST be the first chunk in the datagram. DATA chunks, but it MUST be the first chunk in the datagram.
D) Upon reception of the COOKIE chunk, Endpoint "Z" will reply with D) Upon reception of the COOKIE chunk, Endpoint "Z" will reply with
a COOKIE ACK chunk after building a TCB and marking itself to a COOKIE ACK chunk after building a TCB and marking itself to
the ESTABLISHED state. A COOKIE ACK chunk may be combined with the ESTABLISHED state. A COOKIE ACK chunk may be combined with
any pending DATA chunks (and/or SACK chunks), but the COOKIE ACK any pending DATA chunks (and/or SACK chunks), but the COOKIE ACK
chunk must be the first chunk in the datagram. chunk MUST be the first chunk in the datagram.
IMPLEMENTATION NOTE: an implementation may choose to send the IMPLEMENTATION NOTE: an implementation may choose to send the
Communication Up notification to the SCTP user upon reception Communication Up notification to the SCTP user upon reception
of a valid COOKIE. of a valid COOKIE.
Stewart, et al [Page 37]
E) Upon reception of the COOKIE ACK, endpoint "A" will move from the E) Upon reception of the COOKIE ACK, endpoint "A" will move from the
COOKIE-SENT state to the ESTABLISHED state, stopping the T1-init COOKIE-SENT state to the ESTABLISHED state, stopping the T1-init
timer, and it may also notify its ULP about the successful timer, and it may also notify its ULP about the successful
establishment of the associate with a Communication Up notification establishment of the associate with a Communication Up notification
(see Section 9). (see Section 9).
Note: DATA chunk MUST NOT be carried in the INIT or INIT ACK message. Note: A DATA chunk MUST NOT be carried in the INIT or INIT ACK message.
Note: T1-init timer shall follow the same rules given in Section 5.3. Note: T1-init timer shall follow the same rules given in Section 5.3.
Note: if an endpoint receives an INIT, INIT ACK, or COOKIE chunk but Note: if an endpoint receives an INIT, INIT ACK, or COOKIE chunk but
decides not to establish the new association due to missing mandatory decides not to establish the new association due to missing mandatory
parameters in the received INIT or INIT ACK, invalid parameter values, parameters in the received INIT or INIT ACK, invalid parameter values,
or, lack of local resources, it SHALL respond with an ABORT chunk. It or, lack of local resources, it SHALL respond with an ABORT chunk. It
SHOULD also bundle with the ABORT one or more Operational ERROR chunks SHOULD also specify the cause of abort, such as the type of the
to specify the cause of abort, such as the type(s) of the missing missing mandatory parameters, etc., by either including cause
mandatory parameters, etc. The Verification Tag field in the common parameters or bundling with the ABORT one or more Operational ERROR
header of the outbound abort datagram MUST be set to equal the chunks. The Verification Tag field in the common header of the
Initiate Tag value of the peer. outbound abort datagram MUST be set to equal the Initiate Tag value of
the peer.
Note: After the reception of the first data chunk in an association Note: After the reception of the first data chunk in an association
the receiver MUST immediately respond with a SACK to acknowledge the receiver MUST immediately respond with a SACK to acknowledge
the data chunk, subsequent acknowledgments should be done as
the data chunk, subsequent acknowledgements should be done as
described in section 5.2. described in section 5.2.
Note: When an SCTP endpoint sends an INIT or INIT ACK it SHOULD Note: When an SCTP endpoint sends an INIT or INIT ACK it SHOULD
include all of its transport addresses in the parameter section. This include all of its transport addresses in the parameter section. This
is because it may NOT be possible to control the "sending" address is because it may NOT be possible to control the "sending" address
that a receiver of an SCTP datagram sees. A receiver thus MUST know that a receiver of an SCTP datagram sees. A receiver thus MUST know
every address that may be a source address for a peer SCTP endpoint, every address that may be a source address for a peer SCTP endpoint,
this assures that the inbound SCTP datagram can be matched to the this assures that the inbound SCTP datagram can be matched to the
proper association. proper association.
Note: At the time when the TCB is created, either end MUST set its Note: At the time when the TCB is created, either end MUST set its
internal cumulative TSN acknowledgment point to its peer's Initial TSN internal cumulative TSN acknowledgment point to its peer's Initial TSN
minus one. minus one.
IMPLEMENTATION Note: The IP address and SCTP port(s) are generally
used as the key to find the TCB within an SCTP instance.
4.1.1 Handle Stream Parameters 4.1.1 Handle Stream Parameters
In the INIT and INIT ACK messages, the sender of the message shall In the INIT and INIT ACK messages, the sender of the message shall
indicate the number of outbound streams (OS) it wishes to have in the indicate the number of outbound streams (OS) it wishes to have in the
association, as well as the maximal inbound streams (MIS) it will association, as well as the maximal inbound streams (MIS) it will
accept from the other endpoint. accept from the other endpoint.
After receiving these stream configuration information from the other After receiving these stream configuration information from the other
side, each endpoint shall perform the following check: if the peer's side, each endpoint shall perform the following check: if the peer's
MIS is less than the endpoint's OS, meaning that the peer is incapable MIS is less than the endpoint's OS, meaning that the peer is incapable
of supporting all the outbound streams the endpoint wants to of supporting all the outbound streams the endpoint wants to
configure, the endpoint MUST either settle with MIS outbound streams, configure, the endpoint MUST either settle with MIS outbound streams,
or abort the association and report to its upper layer the resources or abort the association and report to its upper layer the resources
shortage at its peer. shortage at its peer.
Stewart, et al [Page 38]
After the association is initialized, the valid outbound stream After the association is initialized, the valid outbound stream
identifier range for either endpoint shall be 0 to identifier range for either endpoint shall be 0 to
min(local OS, remote MIS)-1. min(local OS, remote MIS)-1.
4.1.2 Handle Address Parameters 4.1.2 Handle Address Parameters
During the association initialization, an endpoint shall use the During the association initialization, an endpoint shall use the
following rules to discover and collect the destination transport following rules to discover and collect the destination transport
address(es) to its peer. address(es) of its peer.
On reception of an INIT or INIT ACK message, the receiver shall record On reception of an INIT or INIT ACK message, the receiver shall record
any transport addresses specified as parameters in the INIT or INIT any transport addresses. The transport address(es) are derived by the
ACK message, and use only these addresses as destination transport combination of SCTP source port (from the common header) and the IP
addresses when sending subsequent datagrams to its peer. If no address parameter(s) carried in the INIT or INIT ACK message. The
destination transport addresses are specified in the INIT or INIT ACK receiver should use only these transport addresses as destination
message, then the source address from which the message arrived should transport addresses when sending subsequent datagrams to its peer. If
be considered as the only destination transport address to use. no IP address parameters are specified in the INIT or INIT ACK
message, then the source IP address from which the message arrives
should be combined with SCTP source port number and be considered as
the only destination transport address to use.
An initial primary destination transport address shall be selected An initial primary destination transport address shall be selected
for either endpoint, using the following rules: for either endpoint, using the following rules:
For the initiator: any valid transport address obtained from the For the initiator: any valid transport address obtained from the
INIT ACK message. If no transport address is specified in the INIT INIT ACK message. If no transport address is specified in the INIT
ACK message, use the source transport address from which the INIT ACK message, use the source transport address from which the INIT
ACK message arrived. ACK message arrived.
For the responder: any valid transport address obtained from the For the responder: any valid transport address obtained from the
skipping to change at page 42, line 23 skipping to change at page 91, line ?
1) create an association TCB using information from both the received 1) create an association TCB using information from both the received
INIT and the outgoing INIT ACK messages, INIT and the outgoing INIT ACK messages,
2) in the TCB, set the creation time to the current time of day, and 2) in the TCB, set the creation time to the current time of day, and
the lifespan to the protocol parameter 'Valid.Cookie.Life', the lifespan to the protocol parameter 'Valid.Cookie.Life',
3) attach a private security key to the TCB and generate a 128-bit MD5 3) attach a private security key to the TCB and generate a 128-bit MD5
signature from the key/TCB combination (see [4] for details on signature from the key/TCB combination (see [4] for details on
MD5), and MD5), and
Stewart, et al [Page 39]
4) generate the Encryption Cookie by combining the TCB and the 4) generate the Encryption Cookie by combining the TCB and the
resultant MD5 signature. resultant MD5 signature.
After sending the INIT ACK with the cookie, the sender SHOULD delete After sending the INIT ACK with the cookie, the sender SHOULD delete
the TCB and any other local resource related to the new association, the TCB and any other local resource related to the new association,
so as to prevent resource attacks. so as to prevent resource attacks.
The private key should be a cryptographic quality random number with The private key should be a cryptographic quality random number with
a sufficient length. Discussion in RFC 1750 [1] can be helpful in a sufficient length. Discussion in RFC 1750 [1] can be helpful in
selection of the key. selection of the key.
skipping to change at page 43, line 18 skipping to change at page 91, line ?
3) compare the creation time stamp in the cookie to the current local 3) compare the creation time stamp in the cookie to the current local
time, if the elapsed time is longer than the lifespan carried in time, if the elapsed time is longer than the lifespan carried in
the cookie, then the datagram, including the COOKIE and the the cookie, then the datagram, including the COOKIE and the
attached user data, SHOULD be discarded and the endpoint MUST attached user data, SHOULD be discarded and the endpoint MUST
transmit a stale cookie operational error to the sending endpoint, transmit a stale cookie operational error to the sending endpoint,
4) if the cookie is valid, create an association to the sender of the 4) if the cookie is valid, create an association to the sender of the
COOKIE message with the information in the TCB data carried in the COOKIE message with the information in the TCB data carried in the
COOKIE, and enter the ESTABLISHED state, COOKIE, and enter the ESTABLISHED state,
5) acknowledge any DATA chunk in the datagram following the rules 5) immediately acknowledge any DATA chunk in the datagram with a SACK
defined in Section 5.2, and, (subsequent datagram acknowledgement should following the rules
defined in Section 5.2), and,
Stewart, et al [Page 40]
6) send a COOKIE ACK chunk to the sender acknowledging reception of 6) send a COOKIE ACK chunk to the sender acknowledging reception of
the cookie. The COOKIE ACK MAY be piggybacked with any outbound the cookie. The COOKIE ACK MAY be piggybacked with any outbound
DATA chunk or SACK chunk. DATA chunk or SACK chunk.
Note that if a COOKIE is received from an endpoint with which the Note that if a COOKIE is received from an endpoint with which the
receiver of the COOKIE has an existing association, the procedures in receiver of the COOKIE has an existing association, the procedures in
section 4.2 should be followed. section 4.2 should be followed.
4.1.6 An Example of Normal Association Establishment 4.1.6 An Example of Normal Association Establishment
skipping to change at page 44, line 18 skipping to change at page 91, line ?
... ...
{app sends 1st user data; strm 0} {app sends 1st user data; strm 0}
DATA [TSN=initial TSN_A DATA [TSN=initial TSN_A
Strm=0,Seq=1 & user data]--\ Strm=0,Seq=1 & user data]--\
(Start T3-rxt timer) \ (Start T3-rxt timer) \
\-> \->
/----- SACK [TSN ACK=init TSN_A,Frag=0] /----- SACK [TSN ACK=init TSN_A,Frag=0]
(Cancel T3-rxt timer) <------/ (Cancel T3-rxt timer) <------/
... ...
Stewart, et al [Page 41]
... ...
{app sends 2 datagrams;strm 0} {app sends 2 datagrams;strm 0}
/---- DATA /---- DATA
/ [TSN=init TSN_Z / [TSN=init TSN_Z
<--/ Strm=0,Seq=1 & user data 1] <--/ Strm=0,Seq=1 & user data 1]
SACK [TSN ACK=init TSN_Z, /---- DATA SACK [TSN ACK=init TSN_Z, /---- DATA
Frag=0] --------\ / [TSN=init TSN_Z +1, Frag=0] --------\ / [TSN=init TSN_Z +1,
\/ Strm=0,Seq=2 & user data 2] \/ Strm=0,Seq=2 & user data 2]
<------/\ <------/\
\ \
skipping to change at page 44, line 41 skipping to change at page 91, line ?
chunks are sent, the same INIT or cookie chunk with the same Initiate chunks are sent, the same INIT or cookie chunk with the same Initiate
Tag (i.e., Tag_A) or cookie shall be retransmitted and the timer Tag (i.e., Tag_A) or cookie shall be retransmitted and the timer
restarted. This shall be repeated Max.Init.Retransmits times before "A" restarted. This shall be repeated Max.Init.Retransmits times before "A"
considers "Z" unreachable and reports the failure to its upper layer considers "Z" unreachable and reports the failure to its upper layer
(and thus the association enters the CLOSED state). When (and thus the association enters the CLOSED state). When
retransmitting the INIT, the endpoint SHALL following the rules retransmitting the INIT, the endpoint SHALL following the rules
defined in 5.3 to determine the proper timer value. defined in 5.3 to determine the proper timer value.
4.2 Handle Duplicate INIT, INIT ACK, COOKIE, and COOKIE ACK 4.2 Handle Duplicate INIT, INIT ACK, COOKIE, and COOKIE ACK
At any time during the life of an association (in one of the possible During the life time of an association (in one of the possible
states) between an endpoint and its peer, one of the setup chunks states), an endpoint may receive from its peer endpoint one of the
may be received from the peer, the receiver shall process such setup chunks (INIT, INIT ACK, COOKIE, and COOKIE ACK). The receiver
a duplicate setup chunk as described in this section. shall treat such a setup chuck as a duplicate and process it as
described in this section.
The following scenarios can cause duplicated chunks: The following scenarios can cause duplicated chunks:
A) The peer has crashed without being detected, and re-started itself A) The peer has crashed without being detected, and re-started itself
and sent out a new INIT Chunk trying to restore the association, and sent out a new INIT Chunk trying to restore the association,
B) Both sides are trying to initialize the association at about the B) Both sides are trying to initialize the association at about the
same time, same time,
C) The chunk is from a staled datagram that was used to establish C) The chunk is from a staled datagram that was used to establish
the present association or a past association which is no longer in the present association or a past association which is no longer in
existence, or existence,
D) The chunk is a false message generated by an attacker. D) The chunk is a false message generated by an attacker, or
E) The peer never received the COOKIE ACK and is retransmitting its
COOKIE.
In case A), the endpoint shall reset the present association and set a In case A), the endpoint shall reset the present association and set a
new association with its peer. Case B) is unique and is discussed in new association with its peer. Case B) is unique and is discussed in
Section 4.2.1. However, in cases C) and D), the endpoint must retain Section 4.2.1. However, in cases C), D) and E), the endpoint must retain
the present association. the present association.
The rules in the following sections shall be applied in order to The rules in the following sections shall be applied in order to
identify and correctly handle these cases. identify and correctly handle these cases.
Stewart, et al [Page 42]
4.2.1 Handle Duplicate INIT in COOKIE-WAIT or COOKIE-SENT State 4.2.1 Handle Duplicate INIT in COOKIE-WAIT or COOKIE-SENT State
This usually indicates an initialization collision, i.e., both This usually indicates an initialization collision, i.e., both
endpoints are attempting at about the same time to establish an endpoints are attempting at about the same time to establish an
association with the other endpoint. association with the other endpoint.
In such a case, each of the two side shall respond to the other side In such a case, each of the two side shall respond to the other side
with an INIT ACK, with the Verification Tag field of the common header with an INIT ACK, with the Verification Tag field of the common header
set to the tag value received from the INIT message, and the Initiate set to the tag value received from the INIT message, and the Initiate
Tag field set to its own tag value (the same tag used in the INIT Tag field set to its own tag value (the same tag used in the INIT
message sent out by itself). Each responder shall also generate a message sent out by itself). Each responder shall also generate a
cookie with the INIT ACK. cookie with the INIT ACK.
After that, no other actions shall be taken by either side, i.e., the After that, no other actions shall be taken by either side, i.e., the
endpoint shall not change its state, and the T1-init timer shall be endpoint shall not change its state, and the T1-init timer shall be
let running. The normal procedures for handling cookies will left running. The normal procedures for handling cookies will
resolve the duplicate INITs to a single association. resolve the duplicate INITs to a single association.
4.2.2 Handle Duplicate INIT in Other States 4.2.2 Handle Duplicate INIT in Other States
Upon reception of the duplicated INIT, the receiver shall generate an Upon reception of the duplicated INIT, the receiver shall generate an
INIT ACK with an Encryption Cookie. INIT ACK with an Encryption Cookie.
In the outbound INIT ACK, the endpoint shall set the Verification Tag In the outbound INIT ACK, the endpoint shall set the Verification Tag
field in the common header to the peer's new tag value (from the field in the common header to the peer's new tag value (from the
duplicated INIT message), and the Initiate Tag field to its own tag duplicated INIT message), and the Initiate Tag field to its own tag
skipping to change at page 46, line 20 skipping to change at page 91, line ?
processing of an old INIT or duplicated INIT message. processing of an old INIT or duplicated INIT message.
4.2.4 Handle Duplicate Cookie 4.2.4 Handle Duplicate Cookie
When a duplicated COOKIE chunk is received in any state for an When a duplicated COOKIE chunk is received in any state for an
existing association the following rules shall be applied: existing association the following rules shall be applied:
1) compute an MD5 signature using the TCB data carried in the cookie 1) compute an MD5 signature using the TCB data carried in the cookie
along with the receiver's private security key, along with the receiver's private security key,
Stewart, et al [Page 43]
2) authenticate the cookie by comparing the computed MD5 signature 2) authenticate the cookie by comparing the computed MD5 signature
against the one carried in the cookie. If this comparison fails, against the one carried in the cookie. If this comparison fails,
the datagram, including the COOKIE and the attached user data, the datagram, including the COOKIE and the attached user data,
should be silently discarded (this is case C or D above). should be silently discarded (this is case C or D above).
3) compare the timestamp in the cookie to the current time, if 3) compare the timestamp in the cookie to the current time, if
the cookie is older than the lifespan carried in the cookie, the cookie is older than the lifespan carried in the cookie,
the datagram, including the COOKIE and the attached user data, the datagram, including the COOKIE and the attached user data,
should be discarded and the endpoint MUST transmit a stale cookie should be discarded and the endpoint MUST transmit a stale cookie
error to the sending endpoint only if the Verification tags of the error to the sending endpoint only if the Verification tags of the
cookie's TCB does NOT match the current tag values in the association cookie's TCB does NOT match the current tag values in the association
(this is case C or D above). (this is case C or D above). If both Verification tags do match
consider the cookie valid (this is case E).
4) If the cookie proves to be valid, unpack the TCB into a 4) If the cookie proves to be valid, unpack the TCB into a
temporary TCB. temporary TCB.
5) If the Verification Tags in the Temporary TCB matches the 5) If the Verification Tags in the Temporary TCB matches the
Verification Tags in the existing TCB, the cookie is a Verification Tags in the existing TCB, the cookie is a
duplicate cookie. A cookie ack should be sent to the peer duplicate cookie. A cookie ack should be sent to the peer
endpoint but NO update should be made to the existing endpoint but NO update should be made to the existing
TCB. TCB.
6) If the the local Verification Tag in the temporary TCB 6) If the the local Verification Tag in the temporary TCB
does not match the local Verification Tag in the existing does not match the local Verification Tag in the existing
TCB, then the cookie is a old stale cookie and does TCB, then the cookie is an old stale cookie and does
not correspond to the existing association (case C above). not correspond to the existing association (case C above).
The datagram should be silently discarded. The datagram should be silently discarded.
7) If the Peers Verification Tag in the temporary TCB does not 7) If the peer's Verification Tag in the temporary TCB does not
match the Peer's Verification Tag in the existing TCB match the peer's Verification Tag in the existing TCB,
then a restart of the peer has occurred (case A above). then a restart of the peer has occurred (case A above).
In such a case, the endpoint should report the restart to its ULP In such a case, the endpoint should report the restart to its ULP
and respond the peer with a COOKIE ACK message. It shall also and respond the peer with a COOKIE ACK message. It shall also
update the Verification Tag, initial TSN, and the destination update the Verification Tag, initial TSN, and the destination
address list of the existing TCB with the information from the address list of the existing TCB with the information from the
temporary TCB. After that the temporary TCB can be discarded. temporary TCB. After that the temporary TCB can be discarded.
Furthermore, all the congestion control parameters (e.g., cwnd, Furthermore, all the congestion control parameters (e.g., cwnd,
ssthresh) related to this peer shall be reset to their initial ssthresh) related to this peer shall be reset to their initial
values (see Section 6.2.1). values (see Section 6.2.1).
IMPLEMENTATION NOTE: It is an implementation decision on how IMPLEMENTATION NOTE: It is an implementation decision on how
to handle any pending datagrams. The implementation may elect to handle any pending datagrams. The implementation may elect
to either A) send all messages back to its upper layer with the to either A) send all messages back to its upper layer with the
restart report, or B) automatically re-queue any datagrams restart report, or B) automatically re-queue any datagrams
pending by marking all of them as never-sent and assigning pending by marking all of them as never-sent and assigning
new TSN's at the time of their initial transmissions based upon new TSN's at the time of their initial transmissions based upon
the updated starting TSN (as defined in section 5.5). the updated starting TSN (as defined in section 5).
Note: The "peer's Verification Tag" is the tag received in the INIT
or INIT ACK chunk.
Stewart, et al [Page 44]
4.2.5 Handle Duplicate COOKIE-ACK. 4.2.5 Handle Duplicate COOKIE-ACK.
At any state other than COOKIE-SENT, an endpoint may receive a At any state other than COOKIE-SENT, an endpoint may receive a
duplicated COOKIE ACK chunk. If so, the chunk should be silently duplicated COOKIE ACK chunk. If so, the chunk should be silently
discarded. discarded.
4.2.6 Handle Stale COOKIE Error 4.2.6 Handle Stale COOKIE Error
A stale cookie error indicates one of a number of possible events: A stale cookie error indicates one of a number of possible events:
A) that the association failed to completely setup before the A) that the association failed to completely setup before the
cookie issued by the sender was processed. cookie issued by the sender was processed.
B) an old cookie was processed after setup completed. B) an old cookie was processed after setup completed.
C) an old cookie is received from someone that the receiver is C) an old cookie is received from someone that the receiver is
not interested in having a association with and the ABORT not interested in having an association with and the ABORT
message was lost. message was lost.
When processing a stale cookie an endpoint should first examine When processing a stale cookie an endpoint should first examine
if an association is in the process of being setup, i.e. the if an association is in the process of being setup, i.e. the
association is in the COOKIE-SENT state. In all cases if association is in the COOKIE-SENT state. In all cases if
the association is NOT in the COOKIE-SENT state, the stale the association is NOT in the COOKIE-SENT state, the stale
cookie message should be silently discarded. cookie message should be silently discarded.
If the association is in the COOKIE-SENT state, the endpoint may elect If the association is in the COOKIE-SENT state, the endpoint may elect
one of the following three alternatives. one of the following three alternatives.
1) Send a new INIT message to the endpoint, to generate a new cookie 1) Send a new INIT message to the endpoint, to generate a new cookie
and re-attempt the setup procedure. and re-attempt the setup procedure.
2) Discard the TCB and report to the upper layer the inability of 2) Discard the TCB and report to the upper layer the inability of
setting-up the association. setting-up the association.
3) Send a new INIT message to the endpoint, adding a cookie 3) Send a new INIT message to the endpoint, adding a cookie
preservative parameter requesting an extentsion on the life time of preservative parameter requesting an extension on the life time of
the cookie. When calculating the time extension, an implementation the cookie. When calculating the time extension, an implementation
SHOULD use the RTT information measured based on the previous SHOULD use the RTT information measured based on the previous
COOKIE / Stale COOKIE message exchange, and should add no more COOKIE / Stale COOKIE message exchange, and should add no more
than 1 second beyond the measured RTT, due to a long cookie life than 1 second beyond the measured RTT, due to a long cookie life
time makes the endpoint more subject to a replay attack. time makes the endpoint more subject to a replay attack.
4.3 Other Initialization Issues 4.3 Other Initialization Issues
4.3.1 Selection of Tag Value 4.3.1 Selection of Tag Value
Initiate Tag values should be selected from the range of 0x1 to Initiate Tag values should be selected from the range of 0x1 to
0xffffffff. It is very important that the Tag value be randomized to 0xffffffff. It is very important that the Tag value be randomized to
help protect against "man in the middle" and "sequence number" attacks. help protect against "man in the middle" and "sequence number" attacks.
It is suggested that RFC 1750 [1] be used for the Tag randomization. It is suggested that RFC 1750 [1] be used for the Tag randomization.
Stewart, et al [Page 45]
Moreover, the tag value used by either endpoint in a given association Moreover, the tag value used by either endpoint in a given association
MUST never be changed during the lifetime of the association. However, MUST never be changed during the lifetime of the association. However,
a new tag value MUST be used each time the endpoint tears-down and a new tag value MUST be used each time the endpoint tears-down and
then re-establishes the association to the same peer. then re-establishes the association to the same peer.
5. User Data Transfer 5. User Data Transfer
For transmission efficiency, SCTP defines mechanisms for bundling of For transmission efficiency, SCTP defines mechanisms for bundling of
small user messages and segmentation of large user messages. small user messages and segmentation of large user messages.
The following diagram depicts the flow of user messages through SCTP. The following diagram depicts the flow of user messages through SCTP.
+--------------------------+ +--------------------------+
| User Messages | | User Messages |
+--------------------------+ +--------------------------+
SCTP user ^ | SCTP user ^ |
==================|==|======================================= ==================|==|=======================================
| v (1) | v (1)
+------------------+ +--------------------+ +------------------+ +--------------------+
| SCTP DATA Chunks | |SCTP Control Chunks | | SCTP DATA Chunks | |SCTP Control Chunks |
skipping to change at page 49, line 44 skipping to change at page 91, line ?
sender into a single SCTP datagram for transmission, as long as sender into a single SCTP datagram for transmission, as long as
the final size of the datagram does not exceed the current path the final size of the datagram does not exceed the current path
MTU. The receiver will de-multiplex the datagram back into MTU. The receiver will de-multiplex the datagram back into
the original chunks. the original chunks.
The segmentation and bundling mechanisms, as detailed in Sections 5.9 The segmentation and bundling mechanisms, as detailed in Sections 5.9
and 5.10, are optional to implement by the data sender, but they MUST and 5.10, are optional to implement by the data sender, but they MUST
be implemented by the data receiver, i.e., an SCTP receiver MUST be be implemented by the data receiver, i.e., an SCTP receiver MUST be
prepared to receive and process bundled or segmented data. prepared to receive and process bundled or segmented data.
Stewart, et al [Page 46]
5.1 Transmission of DATA Chunks 5.1 Transmission of DATA Chunks
The following general rules SHALL be applied by the sender for The following general rules SHALL be applied by the sender for
transmission and/or retransmission of outbound DATA chunks: transmission and/or retransmission of outbound DATA chunks:
A) At any given time, the sender MUST NOT transmit new data onto any A) At any given time, the sender MUST NOT transmit new data onto any
destination transport address if it has rwnd or more octets of data destination transport address if its peer's rwnd indicates that the
outstanding. The outstanding data size is defined as the total size peer has no buffer space (i.e. rwnd is 0, see Section 5.2.1).
of ALL data chunks outstanding.
However, regardless of the value of rwnd (including if it is 0), However, regardless of the value of rwnd (including if it is 0),
the sender can always have ONE data packet in flight to the the sender can always have ONE data packet in flight to the
receiver. This rule allows the sender to probe for a change in rwnd receiver. This rule allows the sender to probe for a change in rwnd
that the sender missed due to the update having been lost in that the sender missed due to the update having been lost in
transmission from the receiver to the sender. transmission from the receiver to the sender.
B) At any given time, the sender MUST NOT transmit new data onto a B) At any given time, the sender MUST NOT transmit new data onto a
given transport address if it has cwnd or more octets of data given transport address if it has cwnd or more octets of data
outstanding on that transport address. outstanding on that transport address.
skipping to change at page 50, line 42 skipping to change at page 91, line ?
chunk, as long as the size of the final SCTP datagram does not exceed chunk, as long as the size of the final SCTP datagram does not exceed
the current MTU. See Section 5.2. the current MTU. See Section 5.2.
IMPLEMENTATION Note: when the window is full (i.e., transmission is IMPLEMENTATION Note: when the window is full (i.e., transmission is
disallowed by Rule A and/or Rule B), the sender MAY still accept disallowed by Rule A and/or Rule B), the sender MAY still accept
send requests from its upper layer, but SHALL transmit no more DATA send requests from its upper layer, but SHALL transmit no more DATA
chunks until some or all of the outstanding DATA chunks are chunks until some or all of the outstanding DATA chunks are
acknowledged and transmission is allowed by Rule A and Rule B acknowledged and transmission is allowed by Rule A and Rule B
again. again.
Whenever a transmission or retransmission is made, if T3-rxt timer is Whenever a transmission or retransmission is made to any address, if
not currently running, the sender MUST start the timer. However, if the T3-rxt timer of that address is not currently running, the sender
the timer is already running, the sender SHALL restart the timer ONLY MUST start that timer. However, if the timer of that address is
IF the earliest (i.e., lowest TSN) outstanding DATA chunk is being already running, the sender SHALL restart the timer ONLY IF the
retransmitted. earliest (i.e., lowest TSN) outstanding DATA chunk sent to that
address is being retransmitted.
Stewart, et al [Page 47]
When starting or restarting the T3-rxt timer, the timer value must be When starting or restarting the T3-rxt timer, the timer value must be
adjusted according to the timer rules defined in Sections 5.3.2, adjusted according to the timer rules defined in Sections 5.3.2,
and 5.3.3. and 5.3.3.
5.2 Acknowledgment on Reception of DATA Chunks 5.2 Acknowledgment on Reception of DATA Chunks
The SCTP receiver MUST always acknowledge the SCTP sender about the The SCTP receiver MUST always acknowledge the SCTP sender about the
reception of each DATA chunk. reception of each DATA chunk.
The guidelines on delayed acknowledgment algorithm specified in The guidelines on delayed acknowledgment algorithm specified in
Section 4.2 of RFC 2581 [3] SHOULD be followed. Specifically, an Section 4.2 of RFC 2581 [3] SHOULD be followed. Specifically, an
acknowledgement SHOULD be generated for at least every second datagram acknowledgment SHOULD be generated for at least every second datagram
received, and SHOULD be generated within 200 ms of the arrival of any received, and SHOULD be generated within 200 ms of the arrival of any
unacknowledged datagram. unacknowledged datagram.
IMPLEMENTATION NOTE: the maximal delay for generating an IMPLEMENTATION NOTE: the maximal delay for generating an
acknowledgement may be configured by the SCTP user, either acknowledgment may be configured by the SCTP user, either
statically or dynamically, in order to meet the specific statically or dynamically, in order to meet the specific
timing requirement of the signaling protocol being carried. timing requirement of the signaling protocol being carried.
Acknowledgments MUST be sent in SACK control chunks. A SACK chunk can Acknowledgments MUST be sent in SACK control chunks. A SACK chunk can
acknowledge the reception of multiple DATA chunks. See Section 2.3.3 acknowledge the reception of multiple DATA chunks. See Section 2.3.3
for SACK chunk format. In particular, the SCTP receiver MUST fill in for SACK chunk format. In particular, the SCTP receiver MUST fill in
the Cumulative TSN ACK field to indicate the latest cumulative TSN the Cumulative TSN ACK field to indicate the latest cumulative TSN
number it has received, and any received segments beyond the number it has received, and any received segments beyond the
Cumulative TSN SHALL also be reported. Cumulative TSN SHALL also be reported.
skipping to change at page 52, line 5 skipping to change at page 91, line ?
DATA [TSN=9,Strm=0,Seq=5] ------------> (ack delayed) DATA [TSN=9,Strm=0,Seq=5] ------------> (ack delayed)
(Start T3-rxt timer) (Start T3-rxt timer)
... ...
{App sends 1 message; strm 1} {App sends 1 message; strm 1}
(bundle SACK with DATA) (bundle SACK with DATA)
/----- SACK [TSN Ack=9,Frag=0] \ /----- SACK [TSN Ack=9,Frag=0] \
/ DATA [TSN=6,Strm=1,Seq=2] / DATA [TSN=6,Strm=1,Seq=2]
(cancel T3-rxt timer) <------/ (Start T3-rxt timer) (cancel T3-rxt timer) <------/ (Start T3-rxt timer)
Stewart, et al [Page 48]
(ack delayed) (ack delayed)
... ...
(send ack) (send ack)
SACK [TSN ACK=6,Frag=0] -------------> (cancel T3-rxt timer) SACK [TSN ACK=6,Frag=0] -------------> (cancel T3-rxt timer)
5.2.1 Tracking Peer's Receive Buffer Space 5.2.1 Tracking Peer's Receive Buffer Space
Whenever a SACK arrives, a new updated a_rwnd arrives with it. This Whenever a SACK arrives, a new updated a_rwnd arrives with it. This
value represents the amount of buffer space the sender of the SACK, at value represents the amount of buffer space the sender of the SACK, at
the time of transmitting the SACK, has left of its total receive the time of transmitting the SACK, has left of its total receive
buffer space (as specified in the INIT/INIT-ACK). After processing the buffer space (as specified in the INIT/INIT-ACK). After processing the
SACK, the receiver of the SACK must use the following rules to SACK, the receiver of the SACK must use the following rules to
re-calculate the congestion control rwnd, using the received a_rwnd re-calculate the congestion control rwnd, using the received a_rwnd
value. value.
A) At the establishment of the association, the endpoint initializes A) At the establishment of the association, the endpoint initializes
the congestion control rwnd to the Advertised Receiver Window the congestion control rwnd to the Advertised Receiver Window
Credit (a_rwnd) the peer specified in the INIT or INIT ACK. Credit (a_rwnd) the peer specified in the INIT or INIT ACK.
B) Any time a new DATA chunk is transmitted to a peer, the endpoint B) Any time a DATA chunk is transmitted to a peer, the endpoint
subtracts the data size of the chunk from the rwnd of that peer. subtracts the data size of the chunk from the rwnd of that peer.
D) Any time a SACK arrives, the endpoint performs the following: C) Any time a SACK arrives, the endpoint performs the following:
If all outstanding TSNs are acknowledged by the SACK, adopt If all outstanding TSNs are acknowledged by the SACK, adopt
the a_rwnd value in the SACK as the new rwnd. the a_rwnd value in the SACK as the new rwnd.
Otherwise, take the value of the current rwnd, and add to it the Otherwise, take the value of the current rwnd, and add to it the
data size of any newly acknowledged TSNs that has its BE bits set data size of any newly acknowledged TSNs that has its BE bits set
to 11, OR that moved the cumulative TSN point forward. Then, set to 11, OR that moved the cumulative TSN point forward. Then, set
the congestion control rwnd to the lesser of the calculated value the congestion control rwnd to the lesser of the calculated value
and the a_rwnd carried in the SACK. and the a_rwnd carried in the SACK.
E) Any time the T3-rxt timer expires causing all outstanding chunks to D) Any time the T3-rxt timer expires on any address, causing all
be marked for retransmission, add all of the data sizes of those outstanding chunks sent to that address to be marked for
chunks to the rwnd. retransmission, add all of the data sizes of those chunks to the rwnd.
E) Any time a DATA chunk is marked for retransmission via the
fast retransmit algorithm (section 6.2.4), add the DATA chunks
size to the rwnd.
5.3 Management of Retransmission Timer 5.3 Management of Retransmission Timer
SCTP uses a retransmission timer T3-rxt to ensure data delivery in the SCTP uses a retransmission timer T3-rxt to ensure data delivery in the
absence of any feedback from the remote data receiver. The duration of absence of any feedback from the remote data receiver. The duration of
this timer is referred to as RTO (retransmission timeout). this timer is referred to as RTO (retransmission timeout).
When the receiver endpoint is multi-homed, the data sender endpoint When the receiver endpoint is multi-homed, the data sender endpoint
will calculate a separate RTO for each different destination transport will calculate a separate RTO for each different destination transport
addresses of the receiver endpoint. addresses of the receiver endpoint.
Stewart, et al [Page 49]
The computation and management of RTO in SCTP follows closely with how The computation and management of RTO in SCTP follows closely with how
TCP manages its retransmission timer. To compute the current RTO, an TCP manages its retransmission timer. To compute the current RTO, an
SCTP sender maintains two state variables per destination transport SCTP sender maintains two state variables per destination transport
address: SRTT (smoothed round-trip time) and RTTVAR (round-trip time address: SRTT (smoothed round-trip time) and RTTVAR (round-trip time
variation). variation).
5.3.1 RTO Calculation 5.3.1 RTO Calculation
The rules governing the computation of SRTT, RTTVAR, and RTO are The rules governing the computation of SRTT, RTTVAR, and RTO are
as follows: as follows:
skipping to change at page 53, line 26 skipping to change at page 91, line ?
C5) Karn's algorithm: RTT measurements MUST NOT be made using C5) Karn's algorithm: RTT measurements MUST NOT be made using
packets that were retransmitted (and thus for which it is packets that were retransmitted (and thus for which it is
ambiguous whether the reply was for the first instance of the ambiguous whether the reply was for the first instance of the
packet or a later instance). packet or a later instance).
C6) Whenever RTO is computed, if it is less than RTO.Min seconds C6) Whenever RTO is computed, if it is less than RTO.Min seconds
then it is rounded up to RTO.Min seconds. The reason for this then it is rounded up to RTO.Min seconds. The reason for this
rule is that RTOs that do not have a high minimum value are rule is that RTOs that do not have a high minimum value are
susceptible to unnecessary timeouts [5]. susceptible to unnecessary timeouts [5].
Stewart, et al [Page 50]
C7) A maximum value may be placed on RTO provided it is at least C7) A maximum value may be placed on RTO provided it is at least
60 seconds. RTO.max seconds.
There is no requirement for the clock granularity G used for computing There is no requirement for the clock granularity G used for computing
RTT measurements and the different state variables, other than RTT measurements and the different state variables, other than
G1) Whenever RTTVAR is computed, if RTTVAR = 0, then adjust G1) Whenever RTTVAR is computed, if RTTVAR = 0, then adjust
RTTVAR <- G. RTTVAR <- G.
Experience [5] has shown that finer clock granularities (<= 100 msec) Experience [5] has shown that finer clock granularities (<= 100 msec)
perform somewhat better than more coarse granularities. perform somewhat better than more coarse granularities.
5.3.2 Retransmission Timer Rules 5.3.2 Retransmission Timer Rules
The rules for managing the retransmission timer are as follows: The rules for managing the retransmission timer are as follows:
R1) Every time a packet containing data is sent (including a R1) Every time a packet containing data is sent to any address (including
retransmission), if the T3-rxt timer is not running, start it a retransmission), if the T3-rxt timer of that address is not
running so that it will expire after RTO seconds. The RTO running, start it running so that it will expire after the RTO of
used here is that obtained after any doubling due to that address. The RTO used here is that obtained after any doubling
previous T3-rxt timer expirations on the coresponding destination due to previous T3-rxt timer expirations on the corresponding
address as discussed in rule E2 below. destination address as discussed in rule E2 below.
R2) Whenever all outstanding data has been acknowledged, turn off the R2) Whenever all outstanding data on an address has been acknowledged,
T3-rxt timer. turn off the T3-rxt timer of that address.
R3) Whenever a SACK is received that acknowledges new data chunks R3) Whenever a SACK is received that acknowledges new data chunks
including the one with the earliest outstanding TSN (i.e., moving including the one with the earliest outstanding TSN on that address,
the cumulative ACK point forward), restart T3-rxt timer with the restart T3-rxt timer of that address with its current RTO.
current RTO.
The following example shows the use of various timer rules (assuming The following example shows the use of various timer rules (assuming
the receiver uses delayed acks). the receiver uses delayed acks).
Endpoint A Endpoint Z Endpoint A Endpoint Z
{App begins to send} {App begins to send}
Data [TSN=7,Strm=0,Seq=3] ------------> (ack delayed) Data [TSN=7,Strm=0,Seq=3] ------------> (ack delayed)
(Start T3-rxt timer) (Start T3-rxt timer)
{App sends 1 message; strm 1} {App sends 1 message; strm 1}
(bundle ack with data) (bundle ack with data)
skipping to change at page 54, line 31 skipping to change at page 91, line ?
/ \ / \
(Re-start T3-rxt timer) <------/ \--> (ack delayed) (Re-start T3-rxt timer) <------/ \--> (ack delayed)
(ack delayed) (ack delayed)
... ...
{send ack} {send ack}
SACK [TSN ACK=6,Frag=0] --------------> (Cancel T3-rxt timer) SACK [TSN ACK=6,Frag=0] --------------> (Cancel T3-rxt timer)
.. ..
(send ack) (send ack)
(Cancel T3-rxt timer) <-------------- SACK [TSN ACK=8,Frag=0] (Cancel T3-rxt timer) <-------------- SACK [TSN ACK=8,Frag=0]
Stewart, et al [Page 51]
5.3.3 Handle T3-rxt Expiration 5.3.3 Handle T3-rxt Expiration
Whenever the retransmission timer T3-rxt expires on a destination Whenever the retransmission timer T3-rxt expires on a destination
address, do the following: address, do the following:
E1) On the destination address where the timer expires, adjust its E1) On the destination address where the timer expires, adjust its
ssthresh with rules defined in Section 6.2.3 and set the ssthresh with rules defined in Section 6.2.3 and set the
cwnd <- MTU. cwnd <- MTU.
E2) On the destination address where the timer expires, set E2) On the destination address where the timer expires, set
RTO <- RTO * 2 ("back off the timer"). The maximum value discussed RTO <- RTO * 2 ("back off the timer"). The maximum value discussed
in rule C7 above may be used to provide an upper bound to this in rule C7 above (RTO.max) may be used to provide an upper bound
doubling operation. to this doubling operation.
E3) Determine how many of the earliest (i.e., lowest TSN) outstanding E3) Determine how many of the earliest (i.e., lowest TSN) outstanding
Data chunks will fit into a single packet, subject to the MTU Data chunks on the address where the T3-rxt has expired that will
constraint for the path corresponding to the destination transport fit into a single packet, subject to the MTU constraint for the
address where the retransmission is being sent to (this may be path corresponding to the destination transport address where the
different from the address where the timer expires [see Section retransmission is being sent to (this may be different from the
5.4]). Call this value K. Bundle and retransmit those K data address where the timer expires [see Section 5.4]). Call this
chunks in a single packet to the address. Note, the sender is value K. Bundle and retransmit those K data chunks in a single
allowed not to bundle, but only retransmit the earliest chunk in packet to the address.
the outbound packet.
E4) Start the retransmission timer T3-rxt on the destination address E4) Start the retransmission timer T3-rxt on the destination address
to where the retransmission is sent, if rule R1 above indicates to to where the retransmission is sent, if rule R1 above indicates to
do so. Note, the RTO to be used for starting T3-rxt should be the do so. Note, the RTO to be used for starting T3-rxt should be the
one of the destination address to where the retransmission is one of the destination address to where the retransmission is
sent, which, when the receiver is multi-homed, may be different sent, which, when the receiver is multi-homed, may be different
from the destination address where the timer expired (see Section from the destination address where the timer expired (see Section
5.4 below). 5.4 below).
Note that after retransmitting, once a new RTT measurement is obtained Note that after retransmitting, once a new RTT measurement is obtained
(which can happen only when new data has been sent and acknowledged, (which can happen only when new data has been sent and acknowledged,
per rule C5, or for a measurement made from a Heartbeat [see Section per rule C5, or for a measurement made from a Heartbeat [see Section
7.3]), the computation in rule C3 is performed, including the 7.3]), the computation in rule C3 is performed, including the
computation of RTO, which may result in "collapsing" RTO back down computation of RTO, which may result in "collapsing" RTO back down
after it has been subject to doubling (rule E2). after it has been subject to doubling (rule E2).
The final rule for managing the retransmission timer concerns failover The final rule for managing the retransmission timer concerns failover
(see Section 5.4.1): (see Section 5.4.1):
F1) Whenever SCTP switches from the current destination transport F1) Whenever SCTP switches from the current destination transport
address to a different one, the current retransmission timer is address to a different one, the current retransmission timers are
left running. As soon as SCTP transmits a packet containing data left running. As soon as SCTP transmits a packet containing data
to the new transport address, restart the timer, using the RTO to the new transport address, start the timer on that transport
value for the path to the new address, if rule R1 indicates to address, using the RTO value of the destination address where
do so. the data is being sent, if rule R1 indicates to do so.
5.4 Multi-homed SCTP Endpoints 5.4 Multi-homed SCTP Endpoints
An SCTP endpoint is considered multi-homed if there are more than one An SCTP endpoint is considered multi-homed if there are more than one
transport addresses that can be used as a destination address to reach transport addresses that can be used as a destination address to reach
that endpoint. that endpoint.
Stewart, et al [Page 52]
Moreover, at the sender side, one of the multiple destination Moreover, at the sender side, one of the multiple destination
addresses of the multi-homed receiver endpoint shall be selected as addresses of the multi-homed receiver endpoint shall be selected as
the primary destination transport address by the UPL (see Sections the primary destination transport address by the UPL (see Sections
4.1.2 and 9.1 for details). 4.1.2 and 9.1 for details).
When the SCTP sender is transmitting to the multi-homed receiver, by When the SCTP sender is transmitting to the multi-homed receiver, by
default the transmission SHOULD always take place on the primary default the transmission SHOULD always take place on the primary
transport address, unless the SCTP user explicitly specifies the transport address, unless the SCTP user explicitly specifies the
destination transport address to use. destination transport address to use.
The acknowledgement SHOULD be transmitted to the same destination The acknowledgment SHOULD be transmitted to the same destination
transport address from which the DATA or control chunk being transport address from which the DATA or control chunk being
acknowledged were received. acknowledged were received.
However, when acknowledging multiple DATA chunks in a single SACK, the However, when acknowledging multiple DATA chunks in a single SACK, the
SACK message may be transmitted to one of the destination transport SACK message may be transmitted to one of the destination transport
addresses from which the DATA or control chunks being acknowledged addresses from which the DATA or control chunks being acknowledged
were received. were received.
Furthermore, when the receiver is multi-homed, the SCTP data sender Furthermore, when the receiver is multi-homed, the SCTP data sender
SHOULD try to retransmit a chunk to an active destination transport SHOULD try to retransmit a chunk to an active destination transport
skipping to change at page 56, line 24 skipping to change at page 91, line ?
Note, retransmissions do not affect the total outstanding data Note, retransmissions do not affect the total outstanding data
count. However, if the data chunk is retransmitted onto a different count. However, if the data chunk is retransmitted onto a different
destination address, both the outstanding data counts on the new destination address, both the outstanding data counts on the new
destination address and the old destination address where the data destination address and the old destination address where the data
chunk was last sent to shall be adjusted accordingly. chunk was last sent to shall be adjusted accordingly.
5.4.1 Failover from Inactive Destination Address 5.4.1 Failover from Inactive Destination Address
Some of the destination transport addresses of a multi-homed SCTP data Some of the destination transport addresses of a multi-homed SCTP data
receiver may become inactive due to either the occurrance of certain receiver may become inactive due to either the occurrence of certain
error conditions (see Section 7.2) or adjustments from SCTP user. error conditions (see Section 7.2) or adjustments from SCTP user.
When there is outbound data to send and the primary destination When there is outbound data to send and the primary destination
transport address becomes inactive (e.g., due to failures), or where transport address becomes inactive (e.g., due to failures), or where
the SCTP user explicitly requests to send data to an inactive the SCTP user explicitly requests to send data to an inactive
destination transport address, before reporting an error to its ULP, destination transport address, before reporting an error to its ULP,
the SCTP sender should try to send the data to an alternate active the SCTP sender should try to send the data to an alternate active
destination transport address if one exists. destination transport address if one exists.
5.5 Stream Identifier and Sequence Number 5.5 Stream Identifier and Stream Sequence Number
Every DATA chunk MUST carry a valid stream identifier. If a DATA chunk Every DATA chunk MUST carry a valid stream identifier. If a DATA chunk
with an invalid stream identifier is received, the receiver shall, with an invalid stream identifier is received, the receiver shall,
after acknowledging the reception of the DATA chunk following the normal after acknowledging the reception of the DATA chunk following the normal
procedure, respond immediately with an ERROR message with cause set to procedure, respond immediately with an ERROR message with cause set to
Invalid Stream Identifier (see Section 2.3.9) and discard the DATA Invalid Stream Identifier (see Section 2.3.9) and discard the DATA
chunk. chunk.
Stewart, et al [Page 53]
The stream sequence number in all the streams shall start from 0x0 The stream sequence number in all the streams shall start from 0x0
when the association is established. Also, when the stream sequence when the association is established. Also, when the stream sequence
number reaches the value 0xffff the next sequence number shall be set number reaches the value 0xffff the next stream sequence number shall
to 0x0. be set to 0x0.
5.6 Ordered and Un-ordered Delivery 5.6 Ordered and Un-ordered Delivery
By default the SCTP receiver shall ensure the DATA chunks within any By default the SCTP receiver shall ensure the DATA chunks within any
given stream be delivered to the upper layer according to the order of given stream be delivered to the upper layer according to the order of
their stream sequence number. If there are DATA chunks arriving out of their stream sequence number. If there are DATA chunks arriving out of
order of their stream sequence number, the receiver MUST hold the order of their stream sequence number, the receiver MUST hold the
received DATA chunks from delivery until they are re-ordered. received DATA chunks from delivery until they are re-ordered.
However, an SCTP sender can indicate that no ordered delivery is However, an SCTP sender can indicate that no ordered delivery is
required on a particular DATA chunk within the stream by setting the U required on a particular DATA chunk within the stream by setting the U
flag of the DATA chunk to 1. flag of the DATA chunk to 1.
In this case, the receiver must bypass the ordering mechanism and In this case, the receiver must bypass the ordering mechanism and
immediately delivery the data to the upper layer (after re-assembly if immediately delivery the data to the upper layer (after re-assembly if
the user data is segmented by the sender). the user data is segmented by the sender).
skipping to change at page 57, line 25 skipping to change at page 91, line ?
This provides an effective way of transmitting "out-of-band" data in a This provides an effective way of transmitting "out-of-band" data in a
given stream. Also, a stream can be used as an "unordered" stream by given stream. Also, a stream can be used as an "unordered" stream by
simply setting the U flag to 1 in all outbound DATA chunks sent simply setting the U flag to 1 in all outbound DATA chunks sent
through that stream. through that stream.
IMPLEMENTATION NOTE: when sending an unordered DATA chunk, an IMPLEMENTATION NOTE: when sending an unordered DATA chunk, an
implementation may choose to place the DATA chunk in an outbound implementation may choose to place the DATA chunk in an outbound
datagram that is at the head of the outbound transmission queue if datagram that is at the head of the outbound transmission queue if
possible. possible.
Note that the 'Sequence Number' field in an un-ordered data chunk has Note that the 'Stream Sequence Number' field in an un-ordered data
no significance; the sender can fill it with arbitrary value, but the chunk has no significance; the sender can fill it with arbitrary
receiver MUST ignore the field. value, but the receiver MUST ignore the field.
5.7 Report Gaps in Received DATA TSNs 5.7 Report Gaps in Received DATA TSNs
Upon the reception of a new DATA chunk, an SCTP receiver shall examine Upon the reception of a new DATA chunk, an SCTP receiver shall examine
the continuity of the TSNs received. If the receiver detects that gaps the continuity of the TSNs received. If the receiver detects that gaps
exist in the received DATA chunk sequence, an SACK with fragment exist in the received DATA chunk sequence, an SACK with fragment
reports shall be sent back immediately. reports shall be sent back immediately.
Based on the segment reports from the SACK, the data sender can Based on the segment reports from the SACK, the data sender can
calculate the missing DATA chunks and make decisions on whether to calculate the missing DATA chunks and make decisions on whether to
retransmit them (see Section 5.3 for details). retransmit them (see Section 5.3 for details).
Multiple gaps can be reported in one single SACK (see Section 2.3.3). Multiple gaps can be reported in one single SACK (see Section 2.3.3).
Note that when the data sender is multi-homed, the SCTP receiver Note that when the data sender is multi-homed, the SCTP receiver
SHOULD always try to send the SACK to the same network from where the SHOULD always try to send the SACK to the same network from where the
last DATA chunk was received. last DATA chunk was received.
Stewart, et al [Page 54]
Upon the reception of the SACK, the data sender SHALL remove all DATA Upon the reception of the SACK, the data sender SHALL remove all DATA
chunks which have been acknowledged by the SACK. The data sender MUST chunks which have been acknowledged by the SACK. The data sender MUST
also treat all the DATA chunks which fall into the gaps between the also treat all the DATA chunks which fall into the gaps between the
fragments reported by the SACK as "missing". The number of "missing" fragments reported by the SACK as "missing". The number of "missing"
reports for each outstanding DATA chunk MUST be recorded by the data reports for each outstanding DATA chunk MUST be recorded by the data
sender in order to make retransmission decision, see Section 6.2.4 for sender in order to make retransmission decision, see Section 6.2.4 for
details. details.
The following example shows the use of SACK to report a gap. The following example shows the use of SACK to report a gap.
skipping to change at page 58, line 28 skipping to change at page 91, line ?
<-----/ <-----/
(remove 6 and 8 from out-queue, (remove 6 and 8 from out-queue,
and strike 7 as "1" missing report) and strike 7 as "1" missing report)
Note: in order to keep the size of the outbound SCTP datagram not to Note: in order to keep the size of the outbound SCTP datagram not to
exceed the current path MTU, the maximal number of fragments that can exceed the current path MTU, the maximal number of fragments that can
be reported within a single SACK chunk is limited. When a single SACK be reported within a single SACK chunk is limited. When a single SACK
can not cover all the fragments needed to be reported due to the MTU can not cover all the fragments needed to be reported due to the MTU
limitation, the endpoint SHALL send only one SACK, reporting the limitation, the endpoint SHALL send only one SACK, reporting the
fragments from the lowest to highest TSNs, within the size limit set fragments from the lowest to highest TSNs, within the size limit set
by the MTU, and leave the remaining highested TSN fragment numbers by the MTU, and leave the remaining highest TSN fragment numbers
unacknowledged. unacknowledged.
5.8 Adler-32 Checksum Calculation 5.8 Adler-32 Checksum Calculation
When sending an SCTP datagram, the sender MUST strengthen the data When sending an SCTP datagram, the sender MUST strengthen the data
integrity of the transmission by including the Adler-32 checksum integrity of the transmission by including the Adler-32 checksum
value calculated on the datagram, as described below. value calculated on the datagram, as described below.
After the datagram is constructed (containing the SCTP common header After the datagram is constructed (containing the SCTP common header
and one or more control or DATA chunks), the sender shall: and one or more control or DATA chunks), the sender shall:
1) fill in the proper Verification Tag in the SCTP common header and 1) fill in the proper Verification Tag in the SCTP common header and
intialize the Adler-32 checksum filed to 0's. initialize the Adler-32 checksum filed to 0's.
2) calculate the Adler-32 checksum of the whole datagram, including the 2) calculate the Adler-32 checksum of the whole datagram, including the
SCTP common header and all the chunks. Refer to Sections 8.2 and 9 SCTP common header and all the chunks. Refer to Sections 8.2 and 9
in [2] for details of the Adler-32 algorithm. And, in [2] for details of the Adler-32 algorithm. And,
3) put the resultant value into the Adler-32 checksum field in the 3) put the resultant value into the Adler-32 checksum field in the
common header, and leave the rest of the bits unchanged. common header, and leave the rest of the bits unchanged.
Stewart, et al [Page 55]
When an SCTP datagram is received, the receiver MUST first check the When an SCTP datagram is received, the receiver MUST first check the
Adler-32 checksum: Adler-32 checksum:
1) store the received Adler-32 checksum value aside, 1) store the received Adler-32 checksum value aside,
2) replace the 32 bits of the Adler-32 checksum field in the received 2) replace the 32 bits of the Adler-32 checksum field in the received
SCTP datagram with all '0's and calculate an Adler-32 checksum SCTP datagram with all '0's and calculate an Adler-32 checksum
value of the whole received datagram. And, value of the whole received datagram. And,
3) verify that the calculated Adler-32 checksum is the same as the 3) verify that the calculated Adler-32 checksum is the same as the
skipping to change at page 60, line 15 skipping to change at page 91, line ?
datagram smaller than or equal to the current MTU, datagram smaller than or equal to the current MTU,
2) the data sender MUST then assign, in sequence, a separate TSN to 2) the data sender MUST then assign, in sequence, a separate TSN to
each of the DATA chunks in the series, each of the DATA chunks in the series,
3) the data sender MUST also set the B/E bits of the first DATA chunk 3) the data sender MUST also set the B/E bits of the first DATA chunk
in the series to '10', the B/E bits of the last DATA chunk in the in the series to '10', the B/E bits of the last DATA chunk in the
series to '01', and the B/E bits of all other DATA chunks in the series to '01', and the B/E bits of all other DATA chunks in the
series to '00'. series to '00'.
Stewart, et al [Page 56]
The data receiver MUST recognize the segmented DATA chunks, by The data receiver MUST recognize the segmented DATA chunks, by
examining the B/E bits in each of the received DATA chunks, and queue examining the B/E bits in each of the received DATA chunks, and queue
the segmented DATA chunks for re-assembly. Then, it shall pass the the segmented DATA chunks for re-assembly. Then, it shall pass the
re-assembled user message to the specific stream for possible re-assembled user message to the specific stream for possible
re-ordering and final dispatching. re-ordering and final dispatching.
Note, if the data receiver runs out of buffer space while still Note, if the data receiver runs out of buffer space while still
waiting for more segments to complete the re-assembly of the message, waiting for more segments to complete the re-assembly of the message,
it should dispatch part of its inbound message through a partial it should dispatch part of its inbound message through a partial
delivery API (see Section 9), freeing some of its receive buffer space delivery API (see Section 9), freeing some of its receive buffer space
skipping to change at page 61, line 4 skipping to change at page 91, line ?
chunks end on a thirty-two-bit word boundary. If the receiver detects chunks end on a thirty-two-bit word boundary. If the receiver detects
a partial chunk, it MUST drop the chunk. a partial chunk, it MUST drop the chunk.
6. Congestion control 6. Congestion control
Congestion control is one of the basic functions in the SCTP protocol. Congestion control is one of the basic functions in the SCTP protocol.
For some applications, it may be likely that adequate resources will For some applications, it may be likely that adequate resources will
be allocated to SCTP traffic to assure prompt delivery of be allocated to SCTP traffic to assure prompt delivery of
time-critical SCTP data - thus it would appear to be unlikely, during time-critical SCTP data - thus it would appear to be unlikely, during
normal operations, that SCTP transmissions encounter severe congestion normal operations, that SCTP transmissions encounter severe congestion
condition. However SCTP must prepare itself for adverse operational condition. However SCTP must prepare itself for adverse operational
conditions, which can develop upon partial network failures or conditions, which can develop upon partial network failures or
unexpected traffic surge. In such situations SCTP must follow correct unexpected traffic surge. In such situations SCTP must follow correct
congestion control steps to recover from congestion quickly in order congestion control steps to recover from congestion quickly in order
to get data delivered as soon as possible. In the absence of network to get data delivered as soon as possible. In the absence of network
congestion, these preventive congestion control algorithms should show congestion, these preventive congestion control algorithms should show
no impact on the protocol performance. no impact on the protocol performance.
Stewart, et al [Page 57]
IMPLEMENTATION NOTE: as far as its specific performance requirements IMPLEMENTATION NOTE: as far as its specific performance requirements
are met, an implementation is always allowed to adopt a more are met, an implementation is always allowed to adopt a more
conservative congestion control algorithm than the one defined conservative congestion control algorithm than the one defined
below. below.
The congestion control algorithms used by SCTP are based on RFC 2581 The congestion control algorithms used by SCTP are based on RFC 2581
[3], "TCP Congestion Control". This section describes how the [3], "TCP Congestion Control". This section describes how the
algorithms defined in RFC 2581 are adopted for use in SCTP. We first algorithms defined in RFC 2581 are adopted for use in SCTP. We first
list differences in protocol designs between TCP and SCTP, and then list differences in protocol designs between TCP and SCTP, and then
describe SCTP's congestion control scheme. The description will use describe SCTP's congestion control scheme. The description will use
skipping to change at page 61, line 49 skipping to change at page 91, line ?
DATA chunk that can be sent within the congestion window, as is the DATA chunk that can be sent within the congestion window, as is the
case in TCP. SCTP SACK leads to different implementations of case in TCP. SCTP SACK leads to different implementations of
fast-retransmit and fast-recovery from that of TCP. fast-retransmit and fast-recovery from that of TCP.
The biggest difference between SCTP and TCP, however, is multi-homing. The biggest difference between SCTP and TCP, however, is multi-homing.
SCTP is designed to establish robust communication associations SCTP is designed to establish robust communication associations
between two end points each of which may be reachable by more than one between two end points each of which may be reachable by more than one
transport address. Potentially different addresses may lead to transport address. Potentially different addresses may lead to
distinguished data paths between the two points, thus ideally one may distinguished data paths between the two points, thus ideally one may
need a separate set of congestion control parameters for each of the need a separate set of congestion control parameters for each of the
paths. The treatment here of congestion control for multihomed paths. The treatment here of congestion control for multi-homed
receivers is new with SCTP and may require refinement in the receivers is new with SCTP and may require refinement in the
future. The current algorithms make the following assumptions: future. The current algorithms make the following assumptions:
o The sender always uses the same destination address until being o The sender always uses the same destination address until being
instructed by the upper layer otherwise. instructed by the upper layer otherwise.
o The sender keeps a separate congestion control parameter set for each o The sender keeps a separate congestion control parameter set for each
of the destination addresses. The parameters should decay if the of the destination addresses. The parameters should decay if the
address is not used for a long enough time period. address is not used for a long enough time period.
o For each of the destination addresses, do slow-start upon the first o For each of the destination addresses, do slow-start upon the first
transmission to that address. transmission to that address.
Stewart, et al [Page 58]
6.2 SCTP Slow-Start and Congestion Avoidance 6.2 SCTP Slow-Start and Congestion Avoidance
The slow start and congestion avoidance algorithms MUST be used by a The slow start and congestion avoidance algorithms MUST be used by a
SCTP sender to control the amount of outstanding data being injected SCTP sender to control the amount of outstanding data being injected
into the network. The congestion control in SCTP is employed in regard into the network. The congestion control in SCTP is employed in regard
to the association, not to an individual stream. In some situations it to the association, not to an individual stream. In some situations it
may be beneficial for an SCTP sender to be more conservative than the may be beneficial for an SCTP sender to be more conservative than the
algorithms allow, however an SCTP sender MUST NOT be more aggressive algorithms allow, however an SCTP sender MUST NOT be more aggressive
than the following algorithms allow. than the following algorithms allow.
skipping to change at page 63, line 10 skipping to change at page 91, line ?
some implementations use the size of the receiver advertised window). some implementations use the size of the receiver advertised window).
o Whenever cwnd is greater than zero, the sender is allowed to have cwnd o Whenever cwnd is greater than zero, the sender is allowed to have cwnd
octets of data outstanding on that transport address. octets of data outstanding on that transport address.
o When cwnd is less than or equal to ssthresh an SCTP sender MUST use o When cwnd is less than or equal to ssthresh an SCTP sender MUST use
the slow start algorithm to increase cwnd (assuming the current the slow start algorithm to increase cwnd (assuming the current
congestion window is being fully utilized). If the incoming SACK congestion window is being fully utilized). If the incoming SACK
advances the cumulative TSN, cwnd MUST be increased by at most the advances the cumulative TSN, cwnd MUST be increased by at most the
lesser of 1) the total size of the previously outstanding DATA lesser of 1) the total size of the previously outstanding DATA
chunk(s) covered by this advancement of the cumulative TSN, and 2) chunk(s) acknowledged, and 2) the destinations path MTU.
the current MTU. This prevents against the ACK-Splitting attack This prevents against the ACK-Splitting attack outlined in [15].
outlined in [15].
Stewart, et al [Page 59]
NOTE: In instances where the data receiver endpoint is multi-homed,
if a SACK arrives at the data sender that advances the
sender's cumulative TSN point, then the data sender should update
its cwnd (or cwnds) apportioned to the destination addresses where
the data was transmitted to. However if the SACK does not advance
the cumulative TSN point, the data sender MUST not adjust the cwnd
of any of the destination addresses.
NOTE: because an SCTP data sender's cwnd is not tied to its NOTE: because an SCTP data sender's cwnd is not tied to its
cumulative TSN point, as duplicate SACKs come in, even though they cumulative TSN point, as duplicate SACKs come in, even though they
may not advance the cumulative TSN point an SCTP sender can still may not advance the cumulative TSN point an SCTP sender can still
use them to clock out new data. That is, the data newly use them to clock out new data. That is, the data newly
acknowledged by the SACK diminishes the amount of data now in acknowledged by the SACK diminishes the amount of data now in
flight to less than cwnd; and so the current, unchanged value of flight to less than cwnd; and so the current, unchanged value of
cwnd now allows new data to be sent. On the other hand, the cwnd now allows new data to be sent. On the other hand, the
increase of cwnd must be tied to the cumulative TSN advancement as increase of cwnd must be tied to the cumulative TSN advancement as
specified above. Otherwise the duplicate SACKs will not only clock specified above. Otherwise the duplicate SACKs will not only clock
skipping to change at page 63, line 55 skipping to change at page 91, line ?
o When partial_bytes_acked is equal or greater than cwnd and before o When partial_bytes_acked is equal or greater than cwnd and before
the arrival of the SACK the sender has cwnd or more octets of data the arrival of the SACK the sender has cwnd or more octets of data
outstanding, increase cwnd by MTU, and reset partial_bytes_acked to outstanding, increase cwnd by MTU, and reset partial_bytes_acked to
(partial_bytes_acked - cwnd). (partial_bytes_acked - cwnd).
o Same as in the slow start, when the sender does not transmit data on o Same as in the slow start, when the sender does not transmit data on
a given transport address, the cwnd of the transport address should a given transport address, the cwnd of the transport address should
be adjusted to max(cwnd / 2, 2*MTU) per RTO. be adjusted to max(cwnd / 2, 2*MTU) per RTO.
Stewart, et al [Page 60]
6.2.3 Congestion Control 6.2.3 Congestion Control
Upon detection of packet losses from SACK, the sender should do the Upon detection of packet losses from SACK reports (see section 6.2.4),
following: the sender should do the following:
ssthresh = max(cwnd/2, 2*MTU) ssthresh = max(cwnd/2, 2*MTU)
cwnd = ssthresh cwnd = ssthresh
Basically, a packet loss causes cwnd to be cut in half. Basically, a packet loss causes cwnd to be cut in half.
When the T3-rxt timer expires, SCTP should perform slow start by: When the T3-rxt timer expires on an address, SCTP should perform
slow start by:
ssthresh = max(cwnd/2, 2*MTU) ssthresh = max(cwnd/2, 2*MTU)
cwnd = 1*MTU cwnd = 1*MTU
and assuring that no more than one data packet will be in flight until and assure that no more than one data packet will be in flight on that
the sender receives acknowledgment for successful delivery. address until the sender receives acknowledgment for successful delivery
of data to that address.
6.2.4 Fast Retransmit on Gap Reports 6.2.4 Fast Retransmit on Gap Reports
In the absence of data losses, a SCTP receiver performs delayed In the absence of data losses, a SCTP receiver performs delayed
acknowledgment. However, whenever a receiver notices a hole in the acknowledgment. However, whenever a receiver notices a hole in the
arriving TSN sequence, it should start sending a SACK back every time arriving TSN sequence, it should start sending a SACK back every time
a packet arrives carrying data. a packet arrives carrying data.
At the sender end, whenever the sender receives a SACK that indicate At the sender end, whenever the sender receives a SACK that indicate
some TSN(s) missing, it SHOULD wait for 3 further miss indications some TSN(s) missing, it SHOULD wait for 3 further miss indications
skipping to change at page 64, line 31 skipping to change at page 91, line ?
2) Adjust the ssthresh and cwnd of the destination address(es) where 2) Adjust the ssthresh and cwnd of the destination address(es) where
the missing data chunks were last sent, according to the formula the missing data chunks were last sent, according to the formula
described in Section 6.2.3. described in Section 6.2.3.
3) Determine how many of the earliest (i.e., lowest TSN) missing Data 3) Determine how many of the earliest (i.e., lowest TSN) missing Data
chunks will fit into a single packet, subject to constraint of the chunks will fit into a single packet, subject to constraint of the
path MTU of the destination transport address to which the packet path MTU of the destination transport address to which the packet
is being sent. Call this value K. Retransmit those K data chunks in is being sent. Call this value K. Retransmit those K data chunks in
a single packet. a single packet.
4) Restart T3-rxt timer ONLY IF the last SACK advanced the cumulative 4) Restart T3-rxt timer ONLY IF the last SACK acknowledged the lowest
TSN point, or we are retransmitting the first outstanding Data outstanding TSN number sent to that address, or we are retransmitting
chunk. the first outstanding Data chunk sent to that address.
Note, before the above adjustments, if the received SACK also Note, before the above adjustments, if the received SACK also
acknowledges new data chunks and advances the cumulative TSN point, acknowledges new data chunks and advances the cumulative TSN point,
the cwnd adjustment rules defined in Sections 6.2.1 and 6.2.2 must the cwnd adjustment rules defined in Sections 6.2.1 and 6.2.2 must
be applied first. be applied first.
Stewart, et al [Page 61]
A straightforward implementation of the above requires that the sender A straightforward implementation of the above requires that the sender
keeps a counter for each TSN hole first reported by a SACK; the keeps a counter for each TSN hole first reported by a SACK; the
counter keeps track of whether 3 subsequent SACKs have reported the counter keeps track of whether 3 subsequent SACKs have reported the
same hole. same hole.
Because cwnd in SCTP indirectly bounds the number of outstanding Because cwnd in SCTP indirectly bounds the number of outstanding
TSN's, the effect of TCP fast-recovery is achieved automatically with TSN's, the effect of TCP fast-recovery is achieved automatically with
no adjustment to the congestion control window size. no adjustment to the congestion control window size.
6.3 Path MTU Discovery 6.3 Path MTU Discovery
RFC 1191 [11] discusses "Path MTU Discovery", whereby a sender RFC 1191 [11] discusses "Path MTU Discovery", whereby a sender
maintains an estimate of the maximum transmission unit (MTU) along a maintains an estimate of the maximum transmission unit (MTU) along a
given Internet path and refrains from sending datagrams along that given Internet path and refrains from sending datagrams along that
path which exceed the MTU, other than occasional attempts to probe for path which exceed the MTU, other than occasional attempts to probe for
a change in the path MTU. RFC 1191 is thorough in its discussion of a change in the path MTU. RFC 1191 is thorough in its discussion of
the MTU discovery mechanism and strategies for determining the current the MTU discovery mechanism and strategies for determining the current
end-to-end MTU setting as well as detecting changes in this value. end-to-end MTU setting as well as detecting changes in this value.
RFC 1981 discusses applying the same mechanisms for IPv6. RFC 1981 [12] discusses applying the same mechanisms for IPv6.
An SCTP sender SHOULD apply these techniques, and SHOULD do so on a An SCTP sender SHOULD apply these techniques, and SHOULD do so on a
per-destination-address basis. per-destination-address basis.
There are 4 ways in which SCTP differs from the description in RFC 1191 There are 4 ways in which SCTP differs from the description in RFC 1191
of applying MTU discovery to TCP: of applying MTU discovery to TCP:
1) SCTP associations can span multiple set of addresses. 1) SCTP associations can span multiple set of addresses.
Per the above comment, an SCTP sender MUST maintain separate Per the above comment, an SCTP sender MUST maintain separate
MTU estimates for each destination address of its peer. MTU estimates for each destination address of its peer.
skipping to change at page 65, line 48 skipping to change at page 91, line ?
discussion in section 6.5 of RFC 1191 applies: when retransmitting discussion in section 6.5 of RFC 1191 applies: when retransmitting
a datagram to a remote address for which the datagram appears a datagram to a remote address for which the datagram appears
too large for the path MTU to that address, the datagram SHOULD too large for the path MTU to that address, the datagram SHOULD
be retransmitted without the DF bit set, allowing it to possibly be retransmitted without the DF bit set, allowing it to possibly
be fragmented. Transmissions of new datagrams MUST have DF set. be fragmented. Transmissions of new datagrams MUST have DF set.
Other than these differences, the discussion of TCP's use of MTU Other than these differences, the discussion of TCP's use of MTU
discovery in RFCs 1191 and 1981 applies to SCTP, too, on a discovery in RFCs 1191 and 1981 applies to SCTP, too, on a
per-destination-address basis. per-destination-address basis.
Stewart, et al [Page 62]
7. Fault Management 7. Fault Management
7.1 Endpoint Failure Detection 7.1 Endpoint Failure Detection
The data sender shall keep a counter on the total number of The data sender shall keep a counter on the total number of
consecutive retransmissions to its peer (including retransmissions to consecutive retransmissions to its peer (including retransmissions to
ALL the destination transport addresses of the peer if it is ALL the destination transport addresses of the peer if it is
multi-homed). multi-homed).
If the value of this counter exceeds the limit indicated in the If the value of this counter exceeds the limit indicated in the
protocol parameter 'Association.Max.Retrans', the data sender shall protocol parameter 'Association.Max.Retrans', the data sender shall
consider the peer endpoint unreachable and shall stop transmitting any consider the peer endpoint unreachable and shall stop transmitting any
more data to it (and thus the association enters the CLOSED state). In more data to it (and thus the association enters the CLOSED state). In
addition, the data sender shall report the failure to the upper layer, addition, the data sender shall report the failure to the upper layer,
and optionally report back all outstanding user data remaining in its and optionally report back all outstanding user data remaining in its
outbound queue. The association is automatically terminated when the outbound queue. The association is automatically terminated when the
peer endpoint becomes unreachable. peer endpoint becomes unreachable.
The counter shall be reset each time a datagram sent to that The counter shall be reset each time a datagram sent to that
skipping to change at page 66, line 15 skipping to change at page 91, line ?
If the value of this counter exceeds the limit indicated in the If the value of this counter exceeds the limit indicated in the
protocol parameter 'Association.Max.Retrans', the data sender shall protocol parameter 'Association.Max.Retrans', the data sender shall
consider the peer endpoint unreachable and shall stop transmitting any consider the peer endpoint unreachable and shall stop transmitting any
more data to it (and thus the association enters the CLOSED state). In more data to it (and thus the association enters the CLOSED state). In
addition, the data sender shall report the failure to the upper layer, addition, the data sender shall report the failure to the upper layer,
and optionally report back all outstanding user data remaining in its and optionally report back all outstanding user data remaining in its
outbound queue. The association is automatically terminated when the outbound queue. The association is automatically terminated when the
peer endpoint becomes unreachable. peer endpoint becomes unreachable.
The counter shall be reset each time a datagram sent to that The counter shall be reset each time a datagram sent to that
destination address is acknowledged by the peer endpoint. destination address is acknowledged by the peer endpoint, or
a HEARTBEAT-ACK is received from the peer endpoint.
7.2 Path Failure Detection 7.2 Path Failure Detection
When the remote endpoint is multi-homed, the data sender should keep a When the remote endpoint is multi-homed, the data sender should keep a
'retrans.count' counter for each of the destination transport 'retrans.count' counter for each of the destination transport
addresses of the remote endpoint. addresses of the remote endpoint.
Each time the data sender retransmits an outstanding datagram, the Each time the T3-rxt timer on any address, or when a HEARTBEAT sent to
'retrans.count' counter of the destination address, to which the an idle address is not acknowledged, the 'retrans.count' counter of
datagram was previously sent, will be incremented. When the value in that destination address will be incremented. When the value in
'retrans.count' exceeds the protocol parameter 'Path.Max.Retrans' of 'retrans.count' exceeds the protocol parameter 'Path.Max.Retrans' of
that destination address, the data sender should mark the destination that destination address, the data sender should mark the destination
transport address as inactive, and a notification SHOULD be transport address as inactive, and a notification SHOULD be sent to
sent to the upper layer. the upper layer.
When an outstanding TSN is acknowledged, the data sender shall clear When an outstanding TSN is acknowledged or a HEARTBEAT sent to that
the 'retrans.count' counter of the destination transport address to address is acknowledged with a HEARTBEAT-ACK, the data sender shall
which the datagram was last sent. Note, when the data receiver is clear the 'retrans.count' counter of the destination transport address
multi-homed and the last sent was a retransmission to an alternate to which the datagram was last sent (or HEARTBEAT was sent). Note,
address of the receiver, there exists an ambiguity as to whether or when the data receiver is multi-homed and the last sent was a
not the acknowledgment should be credited to the address of the last retransmission to an alternate address of the receiver, there exists
sent. However, this ambiguity does not seem to bear any significant an ambiguity as to whether or not the acknowledgment should be
consequence to SCTP behavior. If this ambiguity is undesirable, the credited to the address of the last sent. However, this ambiguity does
data sender may choose not to clear the 'retrans.count' counter if the not seem to bear any significant consequence to SCTP behavior. If this
last sent was a retransmission. ambiguity is undesirable, the data sender may choose not to clear the
'retrans.count' counter if the last sent was a retransmission.
Note, when configuring the SCTP endpoint, the user should avoid Note, when configuring the SCTP endpoint, the user should avoid
having the value of 'Association.Max.Retrans' larger than the having the value of 'Association.Max.Retrans' larger than the
summation of the 'Path.Max.Retrans' of all the destination addresses summation of the 'Path.Max.Retrans' of all the destination addresses
for the remote endpoint. Otherwise, all the destination addresses may for the remote endpoint. Otherwise, all the destination addresses may
become inactive while the endpoint still considers the peer endpoint become inactive while the endpoint still considers the peer endpoint
reachable. When this condition occurs, how the SCTP chooses to function reachable. When this condition occurs, how the SCTP chooses to function
is implemenation specific. is implementation specific.
Stewart, et al [Page 63]
Note, when the primary destination address is marked inactive (due to Note, when the primary destination address is marked inactive (due to
excessive retransmissions, for instance), the sender MAY automatically excessive retransmissions, for instance), the sender MAY automatically
transmit new datagrams to an alternate destination address if one transmit new datagrams to an alternate destination address if one
exists and is active. This is, howerver, an implementation option. exists and is active. This is, however, an implementation option.
7.3 Path Heartbeat 7.3 Path Heartbeat
By default, an SCTP endpoint shall monitor the reachability of the By default, an SCTP endpoint shall monitor the reachability of the
idle destination transport address(es) of its peer by sending idle destination transport address(es) of its peer by sending
HEARTBEAT messages periodically to the destination transport HEARTBEAT messages periodically to the destination transport
address(es). address(es).
A destination transport address is considered "idle" if no new chunk A destination transport address is considered "idle" if no new chunk
which can be used for updating path RTT (usually including first which can be used for updating path RTT (usually including first
skipping to change at page 67, line 34 skipping to change at page 91, line ?
from the received HEARTBEAT message. from the received HEARTBEAT message.
Upon the receipt of the HEARTBEAT ACK, the sender of the HEARTBEAT Upon the receipt of the HEARTBEAT ACK, the sender of the HEARTBEAT
should clear the 'retrans.count' counter of the destination transport should clear the 'retrans.count' counter of the destination transport
address to which the HEARTBEAT was sent, and mark the destination address to which the HEARTBEAT was sent, and mark the destination
transport address as active if it is not so marked. The endpoint may transport address as active if it is not so marked. The endpoint may
optionally report to the upper layer when an inactive destination optionally report to the upper layer when an inactive destination
address is marked as active due to the reception of the latest address is marked as active due to the reception of the latest
HEARTBEAT ACK. HEARTBEAT ACK.
Stewart, et al [Page 64]
The receiver of the HEARTBEAT ACK should also perform an RTT The receiver of the HEARTBEAT ACK should also perform an RTT
measurement for that destination transport address using the time measurement for that destination transport address using the time
value carried in the HEARTBEAT ACK message. value carried in the HEARTBEAT ACK message.
On an idle destination address that is allowed to heartbeat, HEARTBEAT On an idle destination address that is allowed to heartbeat, HEARTBEAT
messages is RECOMMENDED to be sent once per RTO of that destination messages is RECOMMENDED to be sent once per RTO of that destination
address, with jittering of +/- 50%, and exponential back-off if the address, with jittering of +/- 50%, and exponential back-off if the
previous HEARTBEAT is unanswered. previous HEARTBEAT is unanswered.
A primitive is provided for the SCTP user to change the heart A primitive is provided for the SCTP user to change the heart
beat interval and turn on or off the heart beat on a given destination beat interval and turn on or off the heart beat on a given destination
address. Note, the heartbeat interval set by the SCTP user on any of address. Note, the heartbeat interval set by the SCTP user on any of
the idle destination addresses SHOULD be no smaller than the RTO of the idle destination addresses SHOULD be no smaller than the RTO of
that distination address. Separate timers may be used to control the that destination address. Separate timers may be used to control the
heartbeat transmission for different idle destination addresses. heartbeat transmission for different idle destination addresses.
7.4 Handle "Out of the blue" Packets 7.4 Handle "Out of the blue" Packets
An SCTP datagram is called an "out of the blue" (OOTB) datagram if it An SCTP datagram is called an "out of the blue" (OOTB) datagram if it
is correctly formed, i.e., passed the receiver's Adler-32 check (see is correctly formed, i.e., passed the receiver's Adler-32 check (see
Section 5.8), but the receiver is not able to identify the association Section 5.8), but the receiver is not able to identify the association
to which this datagram belongs. to which this datagram belongs.
The receiver of an OOTB datagram MUST do the following: The receiver of an OOTB datagram MUST do the following:
1) check if the OOTB datagram contains an ABORT chunk. If so, the 1) check if the OOTB datagram contains an ABORT chunk. If so, the
receiver MUST silently discarded the OOTB datagram and take no receiver MUST silently discarded the OOTB datagram and take no
futher action. Otherwise, further action. Otherwise,
2) the receiver should respond the sender of the OOTB datagram with an 2) the receiver should respond the sender of the OOTB datagram with an
ABORT. When sending the ABORT, the receiver of the OOTB datagram ABORT. When sending the ABORT, the receiver of the OOTB datagram
MUST fill in the Verification Tag field of the outbound datagram MUST fill in the Verification Tag field of the outbound datagram
with the value found in the Verification Tag field of the OOTB with the value found in the Verification Tag field of the OOTB
datagram. After sending this ABORT, the receiver of the OOTB datagram. After sending this ABORT, the receiver of the OOTB
datagram shall discard the OOTB datagram and take no further datagram shall discard the OOTB datagram and take no further
action. action.
7.5 Verification Tag 7.5 Verification Tag
The Verification Tag rules defined in this section apply when sending The Verification Tag rules defined in this section apply when sending
or receiving SCTP datagrams which do NOT contain an INIT, SHUTDOWN or receiving SCTP datagrams which do NOT contain an INIT, SHUTDOWN
ACK, or ABORT chunk. The rules for sending and receiving SCTP ACK, or ABORT chunk. The rules for sending and receiving SCTP
datagrams containing one of these chunk types are discussed separately datagrams containing one of these chunk types are discussed separately
in Section 7.5.1. in Section 7.5.1.
When sending an SCTP datagram, the sender MUST fill in the When sending an SCTP datagram, the sender MUST fill in the
Verification Tag field of the outbound datagram with the tag value of Verification Tag field of the outbound datagram with the tag value of
the peer endpoint to which this SCTP datagram is destinated. the peer endpoint to which this SCTP datagram is destined.
When receiving an SCTP datagram, the receiver MUST ensure that the When receiving an SCTP datagram, the receiver MUST ensure that the
value in the Verification Tag field of the received SCTP datagram value in the Verification Tag field of the received SCTP datagram
matches its own Tag. If the received tag value does not match the matches its own Tag. If the received tag value does not match the
receiver's own tag value, the receiver shall silently discard the receiver's own tag value, the receiver shall silently discard the
datagram and shall not process it any further. datagram and shall not process it any further.
Stewart, et al [Page 65]
7.5.1 Exceptions in Verification Tag Rules 7.5.1 Exceptions in Verification Tag Rules
A) Rules for datagram carrying INIT: A) Rules for datagram carrying INIT:
- The sender MUST set the Verification Tag of the datagram to 0. - The sender MUST set the Verification Tag of the datagram to 0.
- The receiver, when noticing an incoming SCTP datagram with the - The receiver, when noticing an incoming SCTP datagram with the
Verification Tag set to 0, should continue to process the datagram Verification Tag set to 0, should continue to process the datagram
only if an INIT chunk is present. Otherwise, the receiver MUST only if an INIT chunk is present. Otherwise, the receiver MUST
silently discard the datagram and take no further action. silently discard the datagram and take no further action.
B) Rules for datagram carrying ABORT: B) Rules for datagram carrying ABORT:
- The sender shall always fill in the Verification Tag field of the - The sender shall always fill in the Verification Tag field of the
outbound datagram with the destination endpoint's tag value if it outbound datagram with the destination endpoint's tag value if it
is known. is known.
- If the ABORT is sent in response to an OOTB datagram, the sender - If the ABORT is sent in response to an OOTB datagram, the sender
MUST follow the procedure described in Section 7.4. MUST follow the procedure described in Section 7.4.
- The receiver MUST accept the datagram IF the Verification Tag - The receiver MUST accept the datagram IF the Verification Tag
matches either its own tag, OR the tag of one of its existing matches either its own tag, OR the tag of its peer. Otherwise, the
peers. Otherwise, the receiver MUST silently discard the datagram receiver MUST silently discard the datagram and take no further
and take no further action. action.
C) Rules for datagram carrying SHUTDOWN ACK: C) Rules for datagram carrying SHUTDOWN ACK:
- When sending a SHUTDOWN ACK, the sender is allowed to either use - When sending a SHUTDOWN ACK, the sender is allowed to either use
the destination endpoint's tag or set the Verification Tag field the destination endpoint's tag or set the Verification Tag field
of the outbound datagram to 0. of the outbound datagram to 0.
- The receiver of a SHUTDOWN ACK shall accept the datagram IF the - The receiver of a SHUTDOWN ACK shall accept the datagram IF the
Verification Tag field of the datagram matches its own tag OR is Verification Tag field of the datagram matches its own tag OR is
set to 0. Otherwise, the receiver MUST silently discard the set to 0. Otherwise, the receiver MUST silently discard the
datagram and take no further action. datagram and take no further action. NOTE: the receiver of the
SHUTDOWN ACK MUST ignore the chunk if it is not in the SHUTDOWN
SENT state.
8. Termination of Association 8. Termination of Association
All existing associations should be terminated when an endpoint exits All existing associations should be terminated when an endpoint exits
from service. An association can be terminated by either close or from service. An association can be terminated by either close or
shutdown. shutdown.
8.1 Close of an Association 8.1 Close of an Association
When an endpoint decides to close down an existing association, it When an endpoint decides to close down an existing association, it
shall send an ABORT message to its peer endpoint. The sender MUST fill shall send an ABORT message to its peer endpoint. The sender MUST fill
in the peer's Verification Tag in the outbound datagram and MUST NOT in the peer's Verification Tag in the outbound datagram and MUST NOT
bundle any other chunk with the ABORT. bundle any DATA chunk with the ABORT.
No acknowledgment is required for an ABORT message. In any No acknowledgment is required for an ABORT message. In any
circumstances, an endpoint MUST NOT respond to any received datagram circumstances, an endpoint MUST NOT respond to any received datagram
that contains an ABORT with its own ABORT (also see Section 7.4). that contains an ABORT with its own ABORT (also see Section 7.4).
The receiver shall apply the special Verification Tag check rules The receiver shall apply the special Verification Tag check rules
described in Section 7.5.1 when handling the datagram carrying an described in Section 7.5.1 when handling the datagram carrying an
ABORT. ABORT.
Stewart, et al [Page 66]
After checking the Verification Tag, the peer shall remove the After checking the Verification Tag, the peer shall remove the
association from its record, and shall report the termination to its association from its record, and shall report the termination to its
upper layer. upper layer.
8.2 Shutdown of an Association 8.2 Shutdown of an Association
Using the TERMINATE primitive (see Section 9.1), the upper layer of an Using the TERMINATE primitive (see Section 9.1), the upper layer of an
endpoint in an association can gracefully shutdown the association. endpoint in an association can gracefully shutdown the association.
This will guarantee that all outstanding datagrams from the peer of This will guarantee that all outstanding datagrams from the peer of
the shutdown initiator be delivered before the association the shutdown initiator be delivered before the association
skipping to change at page 69, line 30 skipping to change at page 91, line ?
them for retransmission and start the retransmit procedure as defined them for retransmission and start the retransmit procedure as defined
in Section 5.3. in Section 5.3.
While in Shutdown-sent state, the initiator shall immediately respond While in Shutdown-sent state, the initiator shall immediately respond
to each inbound SCTP datagram containing user data from the peer with to each inbound SCTP datagram containing user data from the peer with
a SACK and restart the T2-shutdown timer. a SACK and restart the T2-shutdown timer.
If there is no more outstanding datagrams, the peer shall send a If there is no more outstanding datagrams, the peer shall send a
SHUTDOWN ACK and then remove all record of the association. SHUTDOWN ACK and then remove all record of the association.
Stewart, et al [Page 67]
Upon the receipt of the SHUTDOWN ACK, the initiator shall stop the Upon the receipt of the SHUTDOWN ACK, the initiator shall stop the
T2-shutdown timer and remove all record of the association. T2-shutdown timer and remove all record of the association.
Note: that it should be the responsibility of the initiator to assure Note: that it should be the responsibility of the initiator to assure
that all the outstanding datagrams on its side have been resolved that all the outstanding datagrams on its side have been resolved
before it initiates the shutdown procedure. before it initiates the shutdown procedure.
Note: an endpoint shall reject any new data request from its upper Note: an endpoint shall reject any new data request from its upper
layer if it is in Shutdown-sent or Shutdown-received state until layer if it is in Shutdown-sent or Shutdown-received state until
completion of the sequence. completion of the sequence.
skipping to change at page 70, line 7 skipping to change at page 91, line ?
and remove all record of the association. and remove all record of the association.
9. Interface with Upper Layer 9. Interface with Upper Layer
The Upper Layer Protocols (ULP) shall request for services by passing The Upper Layer Protocols (ULP) shall request for services by passing
primitives to SCTP and shall receive notifications from SCTP for primitives to SCTP and shall receive notifications from SCTP for
various events. various events.
The primitives and notifications described in this section should be The primitives and notifications described in this section should be
used as a guideline for implementing SCTP. The following functional used as a guideline for implementing SCTP. The following functional
description of ULP interface primitives is, at best, fictional. We description of ULP interface primitives is shown for illustrative
must warn readers that different SCTP implementations may have purposes. We must warn readers that different SCTP implementations may
different ULP interfaces. However, all SCTPs must provide a certain have different ULP interfaces. However, all SCTPs must provide a
minimum set of services to guarantee that all SCTP implementations can certain minimum set of services to guarantee that all SCTP
support the same protocol hierarchy. This section specifies the implementations can support the same protocol hierarchy.
functional interfaces required of all SCTP implementations.
9.1 ULP-to-SCTP 9.1 ULP-to-SCTP
The following sections functionally characterize a ULP/SCTP interface. The following sections functionally characterize a ULP/SCTP interface.
The notation used is similar to most procedure or function calls in The notation used is similar to most procedure or function calls in
high level languages. high level languages.
The ULP primitives described below specify the basic functions the The ULP primitives described below specify the basic functions the
SCTP must perform to support inter-process communication. Individual SCTP must perform to support inter-process communication. Individual
implementations must define their own exact format, and may provide implementations must define their own exact format, and may provide
combinations or subsets of the basic functions in single calls. combinations or subsets of the basic functions in single calls.
A) Initialize A) Initialize
Format: INITIALIZE ([local port], [local eligible address ]) Format: INITIALIZE ([local port], [local eligible address ])
-> local SCTP instance name -> local SCTP instance name
Stewart, et al [Page 68]
This primitive allows SCTP to initialize its internal data structures This primitive allows SCTP to initialize its internal data structures
and allocate necessary resources for setting up its operation and allocate necessary resources for setting up its operation
environment. Note that once SCTP is initialized, ULP can communicate environment. Note that once SCTP is initialized, ULP can communicate
directly with other endpoints without re-invoking this primitive. directly with other endpoints without re-invoking this primitive.
A local SCTP instance name will be returned to the ULP by the SCTP. A local SCTP instance name will be returned to the ULP by the SCTP.
Mandatory attributes: Mandatory attributes:
None. None.
Optional attributes: Optional attributes:
The following types of attributes may be passed along with the The following types of attributes may be passed along with the
primitive: primitive:
o local port - SCTP port number, if ULP wants it to be specified; o local port - SCTP port number, if ULP wants it to be specified;
o local eliglible address - A single address that the local SCTP o local eligible address - A single address that the local SCTP
endpoint should bind. By default all transport interface cards endpoint should bind. By default all transport interface cards
should be used by the local endpoint. should be used by the local endpoint.
IMPLEMENTAION NOTE: if this optional attribute is supported by an IMPLEMENTATION NOTE: if this optional attribute is supported by an
implementation, it will be the responsibility of the implementation implementation, it will be the responsibility of the implementation
to enforce that the IP source address field of any SCTP datagrams to enforce that the IP source address field of any SCTP datagrams
sent out by this endpoint MUST contain the IP addresses sent out by this endpoint MUST contain the IP addresses
indicated in the local eligible address. indicated in the local eligible address.
B) Associate B) Associate
Format: ASSOCIATE(local SCTP instance name, destination transport Format: ASSOCIATE(local SCTP instance name, destination transport
addr, outbound stream count [,destination eligible transport addr addr, outbound stream count)
list])
-> association id [,destination transport addr list] [,outbound stream -> association id [,destination transport addr list] [,outbound stream
count] count]
This primitive allows the upper layer to initiate an association to a This primitive allows the upper layer to initiate an association to a
specific peer endpoint. specific peer endpoint.
The peer endpoint shall be specified by one of the transport addresses The peer endpoint shall be specified by one of the transport addresses
which defines the endpoint (see section 1.1). If the local SCTP which defines the endpoint (see section 1.4). If the local SCTP
instance has not been initialized, the ASSOCIATE is considered an instance has not been initialized, the ASSOCIATE is considered an
error. error.
The set of transport addresses specified in the "destination eligible
transport addr list" shall be used as valid destinations by the
initiating endpoint when sending to the peer endpoint. If this
parameter is not specified, the initiating endpoint will consider ALL
of the transport addresses returned by the INIT ACK message from the
peer endpoint as valid ones. Note, when specified, the "destination
eligible transport addr list" MUST be the same as, or a sub-set of,
the transport addresses returned in the INIT ACK message from the peer
endpoint. Otherwise, it shall be considered as a configuration error.
An association id, which is a local handle to the SCTP association, An association id, which is a local handle to the SCTP association,
will be returned on successful establishment of the association. If will be returned on successful establishment of the association. If
SCTP is not able to open an SCTP association with the peer endpoint, SCTP is not able to open an SCTP association with the peer endpoint,
an error is returned. an error is returned.
Stewart, et al [Page 69]
Other association parameters may be returned, including the complete Other association parameters may be returned, including the complete
destination transport addresses of the peer as well as the outbound destination transport addresses of the peer as well as the outbound
stream count of the local endpoint. One of the transport address from stream count of the local endpoint. One of the transport address from
the returned destination addresses (or the "destination eligible the returned destination addresses will be selected by the local
transport addr list" if provided) will be selected by the local
endpoint as default primary destination address for sending SCTP endpoint as default primary destination address for sending SCTP
datagrams to this peer. The returned "destination transport addr datagrams to this peer. The returned "destination transport addr
list" can be used by the ULP to change the default primary destination list" can be used by the ULP to change the default primary destination
address or to force sending a datagram to a specific transport address. address or to force sending a datagram to a specific transport address.
IMPLEMENTION NOTE: If ASSOCIATE primitive is implemented as a IMPLEMENTATION NOTE: If ASSOCIATE primitive is implemented as a
blocking function call, the ASSOCIATE primitive can return blocking function call, the ASSOCIATE primitive can return
association parameters in addition to the association id upon association parameters in addition to the association id upon
successful establishment. If ASSOCIATE primitive is implemented as a successful establishment. If ASSOCIATE primitive is implemented as a
non-blocking call, only the association id shall be returned and non-blocking call, only the association id shall be returned and
association parameters shall be passed using the COMMUNICATION UP association parameters shall be passed using the COMMUNICATION UP
notification. notification.
Mandatory attributes: Mandatory attributes:
o local SCTP instance name - obtained from the INITIALIZE operation. o local SCTP instance name - obtained from the INITIALIZE operation.
o destination transport addr - specified as one of the transport o destination transport addr - specified as one of the transport
addresses of the peer endpoint with which the association is to be addresses of the peer endpoint with which the association is to be
established. established.
o outbound stream count - the number of outbound streams the ULP o outbound stream count - the number of outbound streams the ULP
would like to open towards this peer endpoint. would like to open towards this peer endpoint.
Optional attributes: Optional attributes:
o destination eligible transport addr list - a list of transport None.
addresses that the local endpoint is allowed to use for sending
datagrams to this peer. By default, all transport addresses
indicated by the peer in its INIT ACK message can be used.
C) Terminate C) Terminate
Format: TERMINATE(association id) Format: TERMINATE(association id)
-> result -> result
Gracefully terminates an association. Any locally queued user data Gracefully terminates an association. Any locally queued user data
will be delivered to the peer. The association will be terminated only will be delivered to the peer. The association will be terminated only
after the peer acknowledges all the messages sent. A success code after the peer acknowledges all the messages sent. A success code
will be returned on successful termination of the association. If will be returned on successful termination of the association. If
skipping to change at page 72, line 47 skipping to change at page 91, line ?
code shall be returned. code shall be returned.
Mandatory attributes: Mandatory attributes:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
Optional attributes: Optional attributes:
None. None.
Stewart, et al [Page 70]
D) Abort D) Abort
Format: ABORT(association id) Format: ABORT(association id [, cause code])
-> result -> result
Ungracefully terminates an association. Any locally queued user data Ungracefully terminates an association. Any locally queued user data
will be discarded and an ABORT message is sent to the peer. A success will be discarded and an ABORT message is sent to the peer. A success
code will be returned on successful abortion of the association. If code will be returned on successful abortion of the association. If
attempting to abort the association results in a failure, an error attempting to abort the association results in a failure, an error
code shall be returned. code shall be returned.
Note: If possible the SCTP should attempt to return all un-acknowledged
data to the upper layer, however this behavior is implementation
dependent.
Mandatory attributes: Mandatory attributes:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
Optional attributes: Optional attributes:
o cause code - reason of the abort to be passed to the peer.
None. None.
E) Send E) Send
Format: SEND(association id, buffer address, byte count [,context] Format: SEND(association id, buffer address, byte count [,context]
[,stream id] [,life time] [,destination transport address] [,un-order [,stream id] [,life time] [,destination transport address] [,un-order
flag] [,no-bundle flag]) flag] [,no-bundle flag])
-> result -> result
This is the main method to send user data via SCTP. This is the main method to send user data via SCTP.
skipping to change at page 73, line 44 skipping to change at page 91, line ?
Optional attributes: Optional attributes:
o context - optional information that will be carried in the o context - optional information that will be carried in the
sending failure notification to the ULP if the transportation of sending failure notification to the ULP if the transportation of
this datagram fails. this datagram fails.
o stream id - to indicate which stream to send the data on. If not o stream id - to indicate which stream to send the data on. If not
specified, stream 0 will be used. specified, stream 0 will be used.
Stewart, et al [Page 71]
o life time - specifies the life time of the user data. The user data o life time - specifies the life time of the user data. The user data
will not be sent by SCTP after the life time expires. This will not be sent by SCTP after the life time expires. This
parameter can be used to avoid efforts to transmit stale parameter can be used to avoid efforts to transmit stale
user messages. SCTP notifies the ULP, if the data cannot be user messages. SCTP notifies the ULP, if the data cannot be
initiated to transport (i.e. sent to the destination via SCTP's initiated to transport (i.e. sent to the destination via SCTP's
send primitive) within the life time variable. However, the send primitive) within the life time variable. However, the
user data will be transmitted if a TSN has been assigned to the user data will be transmitted if a TSN has been assigned to the
user data before the life time expired. user data before the life time expired.
o destination transport address - specified as one of the destination o destination transport address - specified as one of the destination
skipping to change at page 74, line 37 skipping to change at page 91, line ?
Mandatory attributes: Mandatory attributes:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o destination transport address - specified as one of the transport o destination transport address - specified as one of the transport
addresses of the peer endpoint, which should be used as primary addresses of the peer endpoint, which should be used as primary
address for sending datagrams. This overrides the current primary address for sending datagrams. This overrides the current primary
address information maintained by the local SCTP endpoint. address information maintained by the local SCTP endpoint.
Stewart, et al [Page 72]
G) Receive G) Receive
Format: RECEIVE(association id, buffer address, buffer size Format: RECEIVE(association id, buffer address, buffer size
[,stream id]) [,stream id])
-> byte count [,transport address] [,stream id] [,sequence number] -> byte count [,transport address] [,stream id] [,stream sequence number]
[,partial flag] [, delivery number] [,partial flag] [, delivery number]
This primitive shall read the first user message in the SCTP in-queue This primitive shall read the first user message in the SCTP in-queue
to ULP, if there is one available, into the specified buffer. The size to ULP, if there is one available, into the specified buffer. The size
of the message read, in octets, will be returned. It may, depending on of the message read, in octets, will be returned. It may, depending on
the specific implementation, also return other information such as the the specific implementation, also return other information such as the
sender's address, the stream id on which it is received, whether there sender's address, the stream id on which it is received, whether there
are more messages available for retrieval, etc. For ordered messages, are more messages available for retrieval, etc. For ordered messages,
their sequence number may also be returned. their stream sequence number may also be returned.
Depending upon the implementation, if this primitive is invoked when Depending upon the implementation, if this primitive is invoked when
no message is available the implementation should return an indication no message is available the implementation should return an indication
of this condition or should block the invoking process until data does of this condition or should block the invoking process until data does
become available. become available.
Mandatory attributes: Mandatory attributes:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o buffer address - the memory location indicated by the ULP to store o buffer address - the memory location indicated by the ULP to store
the received message. the received message.
o buffer size - the maximum size of data to be received, in octets. o buffer size - the maximum size of data to be received, in octets.
Optional attributes: Optional attributes:
o stream id - to indicate which stream to receive the data on. o stream id - to indicate which stream to receive the data on.
o sequence number - the stream sequence number assigned by the o stream sequence number - the stream sequence number assigned by the
sending SCTP peer. sending SCTP peer.
o partial flag - if this returned flag is set to 1, then this o partial flag - if this returned flag is set to 1, then this
message is a partial delivery of the whole message. When message is a partial delivery of the whole message. When
this flag is set, the stream id and sequence number MUST this flag is set, the stream id and stream sequence number MUST
accompany this receive. When this flag is set to 0, it indicates accompany this receive. When this flag is set to 0, it indicates
that no more deliveries will be received for this sequence number. that no more deliveries will be received for this stream sequence
number.
Stewart, et al [Page 73]
H) Status H) Status
Format: STATUS(association id) Format: STATUS(association id)
-> status data -> status data
This primitive should return a data block containing the following This primitive should return a data block containing the following
information: information:
association connection state, association connection state,
destination transport address list, destination transport address list,
destination transport address reachability state, destination transport address reachability state,
current receiver window size, current receiver window size,
current congestion window sizes, current congestion window sizes,
number of DATA chunks awaiting acknowledgement, number of DATA chunks awaiting acknowledgment,
number of DATA chunks pending receipt, number of DATA chunks pending receipt,
primary destination transport address, primary destination transport address,
SRTT on primary destination address, SRTT on primary destination address,
RTO on primary destination address, RTO on primary destination address,
SRTT and RTO on other destination addresses, etc. SRTT and RTO on other destination addresses, etc.
Mandatory attributes: Mandatory attributes:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
skipping to change at page 76, line 20 skipping to change at page 91, line ?
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o destination transport address - specified as one of the transport o destination transport address - specified as one of the transport
addresses of the peer endpoint. addresses of the peer endpoint.
o new state - the new state of heart beat for this destination o new state - the new state of heart beat for this destination
transport address (either enabled or disabled). transport address (either enabled or disabled).
Optional attributes: Optional attributes:
o interval - if present, indicates the frequence of the heart beat if o interval - if present, indicates the frequency of the heart beat if
this is to enable heart beat on a destination transport this is to enable heart beat on a destination transport
address. Default interval is the RTO of the destination address. address. Default interval is the RTO of the destination address.
Stewart, et al [Page 74]
J) Request HeartBeat J) Request HeartBeat
Format: REQUESTHEARTBEAT(association id, destination transport Format: REQUESTHEARTBEAT(association id, destination transport
address) address)
-> result -> result
Instructs the local endpoint to perform a HeartBeat on the specified Instructs the local endpoint to perform a HeartBeat on the specified
destination transport address of the given association. The returned destination transport address of the given association. The returned
result should indicate whether the transmission of the HEARTBEAT result should indicate whether the transmission of the HEARTBEAT
message to the destination address is successful. message to the destination address is successful.
skipping to change at page 76, line 49 skipping to change at page 91, line ?
o destination transport address - the transport address of the o destination transport address - the transport address of the
association on which a heartbeat should be issued. association on which a heartbeat should be issued.
K) Get SRTT Report K) Get SRTT Report
Format: GETSRTTREPORT(association id, destination transport address) Format: GETSRTTREPORT(association id, destination transport address)
-> srtt result -> srtt result
Instructs the local SCTP to report the current SRTT measurement on the Instructs the local SCTP to report the current SRTT measurement on the
specified destination transport address of the given association. The specified destination transport address of the given association. The
returned result can be an intager containing the most recent SRTT in returned result can be an integer containing the most recent SRTT in
milliseconds. milliseconds.
Mandatory attributes: Mandatory attributes:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o destination transport address - the transport address of the o destination transport address - the transport address of the
association on which the SRTT measurement is to be reported. association on which the SRTT measurement is to be reported.
L) Set Failure Threshould L) Set Failure Threshold
Format: SETFAILURETHRESHOLD(association id, destination transport Format: SETFAILURETHRESHOLD(association id, destination transport
address, failure threshold) address, failure threshold)
-> result -> result
This primitive allows the local SCTP to customize the reachability This primitive allows the local SCTP to customize the reachability
failure detection threshold 'Path.Max.Retrans' for the specified failure detection threshold 'Path.Max.Retrans' for the specified
destination address. destination address.
Mandatory attributes: Mandatory attributes:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o destination transport address - the transport address of the o destination transport address - the transport address of the
association on which the failure detection threshold is to be set. association on which the failure detection threshold is to be set.
o failure threshold - the new value of 'Path.Max.Retrans' for the o failure threshold - the new value of 'Path.Max.Retrans' for the
destination address. destination address.
Stewart, et al [Page 75]
M) Set Protocol Parameters M) Set Protocol Parameters
Format: SETPROTOCOLPARAMETERS(association id, [,destination transport Format: SETPROTOCOLPARAMETERS(association id, [,destination transport
address,] protocol parameter list) address,] protocol parameter list)
-> result -> result
This primitive allows the local SCTP to customize the protocol This primitive allows the local SCTP to customize the protocol
parameters. parameters.
Mandatory attributes: Mandatory attributes:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o protocol parameter list - The specific names and values of the o protocol parameter list - The specific names and values of the
protocol parameters (e.g., RTO.Initial, Association.Max.Retrans protocol parameters (e.g., Association.Max.Retrans [see Section
[see Section 12]) that the SCTP user wishes to customize. 13]) that the SCTP user wishes to customize.
Optional attributes: Optional attributes:
o destination transport address - some of the protocol parameters may o destination transport address - some of the protocol parameters may
be set on a per destination transport address basis. be set on a per destination transport address basis.
9.2 SCTP-to-ULP 9.2 SCTP-to-ULP
It is assumed that the operating system or application environment It is assumed that the operating system or application environment
provides a means for the SCTP to asynchronously signal the ULP provides a means for the SCTP to asynchronously signal the ULP
skipping to change at page 78, line 4 skipping to change at page 91, line ?
Optional attributes: Optional attributes:
o destination transport address - some of the protocol parameters may o destination transport address - some of the protocol parameters may
be set on a per destination transport address basis. be set on a per destination transport address basis.
9.2 SCTP-to-ULP 9.2 SCTP-to-ULP
It is assumed that the operating system or application environment It is assumed that the operating system or application environment
provides a means for the SCTP to asynchronously signal the ULP provides a means for the SCTP to asynchronously signal the ULP
process. When SCTP does signal an ULP process, certain information is process. When SCTP does signal an ULP process, certain information is
passed to the ULP. passed to the ULP.
IMPLEMENTATION NOTE: in some cases this may be done through a
seperate socket or error channel.
A) DATA ARRIVE notification A) DATA ARRIVE notification
SCTP shall invoke this notification on the ULP when a user message is SCTP shall invoke this notification on the ULP when a user message is
successfully received and ready for retrieval. successfully received and ready for retrieval.
The following may be optionally be passed with the notification: The following may be optionally be passed with the notification:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o stream id - to indicate which stream the data is received on. o stream id - to indicate which stream the data is received on.
Stewart, et al [Page 76]
B) SEND FAILURE notification B) SEND FAILURE notification
If a message can not be delivered SCTP shall invoke this notification If a message can not be delivered SCTP shall invoke this notification
on the ULP. on the ULP.
The following may be optionally be passed with the notification: The following may be optionally be passed with the notification:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o data - the location ULP can find the un-delivered message. o data - the location ULP can find the un-delivered message.
skipping to change at page 79, line 20 skipping to change at page 91, line ?
The following shall be passed with the notification: The following shall be passed with the notification:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o status - This indicates what type of event that has occurred o status - This indicates what type of event that has occurred
o destination transport address list - the complete set of transport o destination transport address list - the complete set of transport
addresses of the peer addresses of the peer
Stewart, et al [Page 77]
o outbound stream count - the maximum number of streams allowed to be o outbound stream count - the maximum number of streams allowed to be
used in this association by the ULP used in this association by the ULP
o inbound stream count - the number of streams the peer endpoint
has requested with this association (this may not be the same
number has 'outbound stream count').
E) COMMUNICATION LOST notification E) COMMUNICATION LOST notification
When SCTP loses communication to an endpoint completely or detects When SCTP loses communication to an endpoint completely or detects
that the endpoint has performed an abort or graceful shutdown that the endpoint has performed an abort or graceful shutdown
operation, it shall invoke this notification on the ULP. operation, it shall invoke this notification on the ULP.
The following shall be passed with the notification: The following shall be passed with the notification:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
skipping to change at page 80, line 5 skipping to change at page 91, line ?
When SCTP receives an ERROR chunk from its peer and decides to notify When SCTP receives an ERROR chunk from its peer and decides to notify
its ULP, it can invoke this notification on the ULP. its ULP, it can invoke this notification on the ULP.
The following can be passed with the notification: The following can be passed with the notification:
o association id - local handle to the SCTP association o association id - local handle to the SCTP association
o error info - this indicates the type of error and optionally some o error info - this indicates the type of error and optionally some
additional information received through the ERROR chunk. additional information received through the ERROR chunk.
Stewart, et al [Page 78]
10. Security Considerations 10. Security Considerations
10.1 Security Objectives 10.1 Security Objectives
As a common transport protocol designed to reliably carry time- As a common transport protocol designed to reliably carry time-
sensitive user messages, such as billing or signalling messages for sensitive user messages, such as billing or signaling messages for
telephony services, between two networked endpoints, SCTP has the telephony services, between two networked endpoints, SCTP has the
following security objectives. following security objectives.
- availability of reliable and timely data transport services - availability of reliable and timely data transport services
- integrity of the user-to-user information carried by SCTP - integrity of the user-to-user information carried by SCTP
10.2 SCTP Responses To Potential Threats 10.2 SCTP Responses To Potential Threats
It is clear that SCTP may potentially be used in a wide variety of It is clear that SCTP may potentially be used in a wide variety of
risk situations. It is important for operator(s) of the systems risk situations. It is important for operator(s) of the systems
concerned to analyze their particular situations and decide on the concerned to analyze their particular situations and decide on the
appropriate counter-measures. appropriate counter-measures.
skipping to change at page 81, line 4 skipping to change at page 91, line ?
separation of services. separation of services.
10.2.2 Protecting against Data Corruption in the Network 10.2.2 Protecting against Data Corruption in the Network
Where the risk of undetected errors in datagrams delivered by the Where the risk of undetected errors in datagrams delivered by the
lower layer transport services is considered to be too great, lower layer transport services is considered to be too great,
additional checksum protection may be required. The question is additional checksum protection may be required. The question is
whether this is appropriately provided as an SCTP service because it whether this is appropriately provided as an SCTP service because it
is needed by most potential users of SCTP, or whether instead it is needed by most potential users of SCTP, or whether instead it
should be provided by the SCTP user application. (The SCTP protocol should be provided by the SCTP user application. (The SCTP protocol
overhead, as opposed to the signaling payload, is protected
overhead, as opposed to the signalling payload, is protected
adequately by the Adler-32 checksum and measures taken in SCTP to prevent adequately by the Adler-32 checksum and measures taken in SCTP to prevent
replay attacks and masquerade.) In any event, the checksum must be replay attacks and masquerade.) In any event, the checksum must be
specifically designed to ensure that it detects the errors left specifically designed to ensure that it detects the errors left
behind by the Adler-32 checksum. behind by the Adler-32 checksum.
Stewart, et al [Page 79]
10.2.3 Protecting Confidentiality 10.2.3 Protecting Confidentiality
In most cases, the risk of breach of confidentiality applies to the In most cases, the risk of breach of confidentiality applies to the
signalling data payload, not to the SCTP or lower-layer protocol signaling data payload, not to the SCTP or lower-layer protocol
overheads. If that is true, encryption of the SCTP user data only overheads. If that is true, encryption of the SCTP user data only
may be considered. As with the supplementary checksum service, user may be considered. As with the supplementary checksum service, user
data encryption may be performed by the SCTP user application. data encryption may be performed by the SCTP user application.
Particularly for mobile users, the requirement for confidentiality Particularly for mobile users, the requirement for confidentiality
may include the masking of IP addresses and ports. In this case may include the masking of IP addresses and ports. In this case
IPSEC ESP should be used instead of application-level encryption. IPSEC ESP should be used instead of application-level encryption.
Similarly, where other reasons prompt the use of the IPSEC ESP Similarly, where other reasons prompt the use of the IPSEC ESP
service, application-level encryption is unnecessary. It will be up service, application-level encryption is unnecessary. It will be up
to the SCTP system operators to configure the application to the SCTP system operators to configure the application
skipping to change at page 81, line 46 skipping to change at page 91, line ?
A blind attack is one where the attacker is unable to intercept or A blind attack is one where the attacker is unable to intercept or
otherwise see the content of data flows passing to and from the otherwise see the content of data flows passing to and from the
target SCTP node where it is not a party to the association. Blind target SCTP node where it is not a party to the association. Blind
denial of service attacks may take the form of flooding, masquerade, denial of service attacks may take the form of flooding, masquerade,
or improper monopolization of services. or improper monopolization of services.
10.2.4.1 Flooding 10.2.4.1 Flooding
The objective of flooding is to cause loss of service and incorrect The objective of flooding is to cause loss of service and incorrect
behaviour at target systems through resource exhaustion, interference behavior at target systems through resource exhaustion, interference
with legitimate transactions, and exploitation of buffer-related with legitimate transactions, and exploitation of buffer-related
software bugs. Flooding may be directed either at the SCTP node or at software bugs. Flooding may be directed either at the SCTP node or at
resources in the intervening IP Access Links or the Internetwork. resources in the intervening IP Access Links or the Internetwork.
Where the latter entities are the target, flooding will manifest Where the latter entities are the target, flooding will manifest
itself as loss of network services, including potentially the breach itself as loss of network services, including potentially the breach
of any firewalls in place. of any firewalls in place.
In general, protection against flooding begins at the equipment In general, protection against flooding begins at the equipment
design level, where it includes measures such as: design level, where it includes measures such as:
- avoiding commitment of limited resources before determining that - avoiding commitment of limited resources before determining that
the request for service is legitimate the request for service is legitimate
- giving priority to completion of processing in progress over the - giving priority to completion of processing in progress over the
acceptance of new work acceptance of new work
- identification and removal of duplicate or stale queued requests - identification and removal of duplicate or stale queued requests
for service. for service.
Stewart, et al [Page 80]
Network equipment should be capable of generating an alarm and log Network equipment should be capable of generating an alarm and log
if a suspicious increase in traffic occurs. The log should provide if a suspicious increase in traffic occurs. The log should provide
information such as the identity of the incoming link and source information such as the identity of the incoming link and source
address(es) used which will help the network or SCTP system operator address(es) used which will help the network or SCTP system operator
to take protective measures. Procedures should be in place for the to take protective measures. Procedures should be in place for the
operator to act on such alarms if a clear pattern of abuse emerges. operator to act on such alarms if a clear pattern of abuse emerges.
The design of SCTP is resistant to flooding attacks, particularly in The design of SCTP is resistant to flooding attacks, particularly in
its use of a four-way start-up handshake, its use of a cookie to its use of a four-way start-up handshake, its use of a cookie to
defer commitment of resources at the responding SCTP node until the defer commitment of resources at the responding SCTP node until the
skipping to change at page 83, line 23 skipping to change at page 91, line ?
10.2.4.3 Improper Monopolization of Services 10.2.4.3 Improper Monopolization of Services
Attacks under this heading are performed openly and legitimately by Attacks under this heading are performed openly and legitimately by
the attacker. They are directed against fellow users of the target the attacker. They are directed against fellow users of the target
SCTP node or of the shared resources between the attacker and the SCTP node or of the shared resources between the attacker and the
target node. Possible attacks include the opening of a large number target node. Possible attacks include the opening of a large number
of associations between the attacker's node and the target, or of associations between the attacker's node and the target, or
transfer of large volumes of information within a legitimately- transfer of large volumes of information within a legitimately-
established association. established association.
Stewart, et al [Page 81]
Such attacks take advantage of policy deficiencies at the target Such attacks take advantage of policy deficiencies at the target
SCTP node. Defense begins with a contractual prohibition of SCTP node. Defense begins with a contractual prohibition of
behaviour directed to denial of service to others. Policy limits behavior directed to denial of service to others. Policy limits
should be placed on the number of associations per adjoining SCTP should be placed on the number of associations per adjoining SCTP
node. SCTP user applications should be capable of detecting large node. SCTP user applications should be capable of detecting large
volumes of illegitimate or "no-op" messages within a given volumes of illegitimate or "no-op" messages within a given
association and either logging or terminating the association as a association and either logging or terminating the association as a
result, based on local policy. result, based on local policy.
10.3 Protection against Fraud and Repudiation 10.3 Protection against Fraud and Repudiation
The objective of fraud is to obtain services without authorization The objective of fraud is to obtain services without authorization
and specifically without paying for them. In order to achieve this and specifically without paying for them. In order to achieve this
skipping to change at page 84, line 22 skipping to change at page 91, line ?
such attacks is seen to exist, or where possible repudiation is an such attacks is seen to exist, or where possible repudiation is an
issue, the use of the IPSEC AH service is recommended to ensure both issue, the use of the IPSEC AH service is recommended to ensure both
the integrity and the authenticity of the messages passed. the integrity and the authenticity of the messages passed.
SCTP also provides no protection against attacks originating at or SCTP also provides no protection against attacks originating at or
beyond the SCTP node and taking place within the context of an beyond the SCTP node and taking place within the context of an
existing association. Prevention of such attacks should be covered existing association. Prevention of such attacks should be covered
by appropriate security policies at the host site, as discussed in by appropriate security policies at the host site, as discussed in
section 10.2.1. section 10.2.1.
11. IANA Consideration Stewart, et al [Page 82]
11. Recommended Transmission Control Block (TCB) Parameters
This section details a recommended set of parameters that should
be contained within the TCB for an implementation. This section is
for illustrative purposes and should not be deemed has requirements
on an implementation NOR as an exhaustive list of all parameters
inside an SCTP TCB. Each implemenation may need its own additional
parameters to optimize their implemenation.
11.1 Parameters necessary for the SCTP instance
Associations - A list of current associations and mappings to the
data consumers for each association. This may be in
the form of a hash table or other implementation dependent
structure. The data consumers may be process identification
information such as file descriptors, named pipe pointer, or
table pointers dependent on how SCTP is implemented.
Secret Key - A secret key used by this endpoint to sign all cookies. This
SHOULD be a cryptographic quality random number with
a sufficient length. Discussion in RFC 1750 [1] can be
helpful in selection of the key.
Address List - The list of IP addresses that this instance has bound. This
information is passed to ones peer('s) in INIT and INIT-ACK
messages.
SCTP Port - The local SCTP port number the endpoint is bound to.
11.2 Parameters necessary per association (i.e. the TCB)
State - A state variable indicating what state the association is
in, i.e . COOKIE_WAIT, COOKIE_SENT, ESTABLISHED,
SHUTDOWN_PENDING, SHUTDOWN_SENT, SHUTDOWN_RECEIVED.
Note: No "CLOSED" state is illustrated since if a
association is "CLOSED" its TCB SHOULD be removed.
Peer Transport
Address List - A list of SCTP transport addresses that the peer is
bound to. This information is derived from the INIT or
INIT-ACK and is used to associate an inbound datagram
with a given association. Normally this information is
hashed or keyed for quick lookup and access of the TCB.
Primary
Destination - This is the current primary destination transport
address of the peer endpoint.
Overall
Error Count - The overall association error count.
Overall Error
Threshold - The threshold for this association that if the Overall
Error Count reaches will cause this association to be
torn down.
Stewart, et al [Page 83]
Per Transport
Address Data - For each destination transport address in the peer's
address list derived from the INIT or INIT ACK message,
a number of data elements needs to be maintained
including:
- Error count - The current error count for this
destination.
- Error Threshold - Current error threshold for
this destination i.e. what value
marks the destination down if
Error count reaches this value.
- cwnd - The current congestion window.
- ssthresh - The current ssthresh value.
- RTO - The current retransmission timeout vaule.
- SRTT - The current smoothed round trip time.
- RTTVAR - The current RTT variation.
- partial_bytes_acked - The tracking method for
increase of cwnd when in
congestion avoidance mode
(see section 6.2.2)
- state - The current state of this destionation,
i.e. DOWN, UP, ALLOW-HB, NO-HEARTBEAT,
etc.
- P-MTU - The current known path MTU.
- Per Destination Timer - A timer used by each
destination.
- RTO-Pending - A flag used to track if one of
the datagrams sent to this address
is currently being used to compute
a RTT. If this flag is 0, the next
datagram sent to this destination
should be used to compute a RTT and
this flag should be set. Every time
the RTT calcualtion completes (i.e.
the datagram is SACK'd) clear this
flag.
- last-timeused - The time this destination was
last sent to. This can be used
to determine if a HEARTBEAT is
needed.
Peer Verification
Tag - Tag value to be sent in every datagram and is received
in the INIT or INIT ACK message.
My Verification
Tag - Tag expected in every inbound datagram and sent in the
INIT or INIT ACK message.
Peer Rwnd - Current calculated value of the peer's rwnd.
Next TSN - My next TSN number I will assign. This is sent in
the INIT or INIT-ACK message to the peer and
incremented each time a DATA chunk is assigned a
TSN (normally just prior to transmit or during
segmentation).
Stewart, et al [Page 84]
Last Rcvd TSN - This is the last TSN I received and is the
current cumulative TSN point. This value is
set initially by taking the peers initial TSN,
received in the INIT or INIT-ACK message, and
subtracting one from it.
Mapping Array - An array of bits or bytes indicating which out of
order TSN's have been received (relative to the
cumulative TSN i.e. Last Rcvd TSN). If no GAP's exist,
i.e. no out of order messages have been received,
this array will be set to all zero. This structure
may be in the form of a circular buffer or bit array.
Ack State - This flag indicates if the next received datagram
is to be responded to with a SACK. This is initialized
to 0, when a datagram is received it is incremented.
If this value reaches 2, a SACK is sent and the value
is reset to 0. Note: this is used only when no datagrams
are received out of order, when DATA chunks are out
of order SACK's are not delayed (see Section 5).
Out Queue - A queue of outbound datagrams.
In Queue - A queue of inbound datagrams.
Reasm Queue - A re-assembly queue.
Inbound
Streams - An array of structures to track the inbound streams.
Normally including the next sequence number expected
and possibly the stream number.
Outbound
Streams - An array of structures to track the outbound streams.
Normally including the next sequence number to
be sent on the stream.
Stewart, et al [Page 85]
12. IANA Consideration
This protocol will require port reservation like TCP for the use of This protocol will require port reservation like TCP for the use of
"well known" servers within the Internet. It is suggested that all "well known" servers within the Internet. It is suggested that all
current TCP ports should be automatically reserved in the SCTP port current TCP ports should be automatically reserved in the SCTP port
address space. New requests should follow IANA's current mechanisms address space. New requests should follow IANA's current mechanisms
for TCP. for TCP.
This protocol may also be extended through IANA in three ways: This protocol may also be extended through IANA in three ways:
-- through definition of additional chunk types, -- through definition of additional chunk types,
-- through definition of additional parameter types, or -- through definition of additional parameter types, or
-- through definition of additional cause codes within Operation -- through definition of additional cause codes within Operation
Error chunks Error chunks
In the case where a particular ULP using SCTP desires to have its own In the case where a particular ULP using SCTP desires to have its own
ports, the ULP should be responsible for registering with IANA for ports, the ULP should be responsible for registering with IANA for
getting its ports assigned. getting its ports assigned.
11.1 IETF-defined Chunk Extension 12.1 IETF-defined Chunk Extension
The appropriate use of specific chunk types is an integral part of the The appropriate use of specific chunk types is an integral part of the
SCTP protocol. In consequence, the intention is that new IETF-defined SCTP protocol. In consequence, the intention is that new IETF-defined
chunk types MUST be supported by standards-track RFC documentation. chunk types MUST be supported by standards-track RFC documentation.
As a transitional step, a new chunk type MAY be introduced in an As a transitional step, a new chunk type MAY be introduced in an
Experimental RFC. Chunk type codes MUST remain permanently associated Experimental RFC. Chunk type codes MUST remain permanently associated
with the original documentation on the basis of which they were with the original documentation on the basis of which they were
allocated. Thus if the RFC supporting a given chunk type is allocated. Thus if the RFC supporting a given chunk type is
deprecated in favour of a new document, the corresponding chunk type deprecated in favor of a new document, the corresponding chunk type
code value is also deprecated and a new code value is allocated in code value is also deprecated and a new code value is allocated in
association with the replacement document. association with the replacement document.
The documentation for a new chunk code type must include the following The documentation for a new chunk code type must include the following
information: information:
(a) a long and short name for the new chunk type; (a) a long and short name for the new chunk type;
(b) a detailed description of the structure of the chunk, which MUST (b) a detailed description of the structure of the chunk, which MUST
conform to the basic structure defined in section 2.2; conform to the basic structure defined in section 2.2;
(c) a detailed definition and description of intended use of each field (c) a detailed definition and description of intended use of each field
within the chunk, including the chunk flags if any; within the chunk, including the chunk flags if any;
(d) a detailed procedural description of the use of the new chunk type (d) a detailed procedural description of the use of the new chunk type
within the operation of the protocol. within the operation of the protocol.
If the primary numbering space reserved for IETF use (0x00 to 0xFD) is If the primary numbering space reserved for IETF use (0x00 to 0xFD) is
exhausted, new codes shall subsequently be allocated in the extension exhausted, new codes shall subsequently be allocated in the extension
range 0x0000 through 0xFFFF. Chunks allocated in this range MUST range 0x0000 through 0xFFFF. Chunks allocated in this range MUST
conform to the following structure: conform to the following structure:
Stewart, et al [Page 86]
First word (32 bits): First word (32 bits):
as shown in section 2.2, with chunk type code equal to 0xFF. as shown in section 2.2, with chunk type code equal to 0xFF.
Second word: Second word:
first octet MUST be all 1's (0xFF). Next octet MUST be all 0's first octet MUST be all 1's (0xFF). Next octet MUST be all 0's
(0x00). Final two octets contain the allocated extension code value. (0x00). Final two octets contain the allocated extension code value.
0 1 2 3 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 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
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|1 1 1 1 1 1 1 1|Chunk Flags | Chunk Length | |1 1 1 1 1 1 1 1|Chunk Flags | Chunk Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|1 1 1 1 1 1 1 1|0 0 0 0 0 0 0 0| Extension Type Code | |1 1 1 1 1 1 1 1|0 0 0 0 0 0 0 0| Extension Type Code |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
\ \ \ \
/ Value / / Value /
\ \ \ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
11.2 IETF-defined Chunk Parameter Extension 12.2 IETF-defined Chunk Parameter Extension
The allocation of a new chunk parameter type code from the IETF The allocation of a new chunk parameter type code from the IETF
numbering space MUST be supported by RFC documentation. As with chunk numbering space MUST be supported by RFC documentation. As with chunk
type codes, parameter type codes are uniquely associated with their type codes, parameter type codes are uniquely associated with their
supporting document and MUST be replaced if new documentation is supporting document and MUST be replaced if new documentation is
provided. This documentation may be Informational, Experimental, or provided. This documentation may be Informational, Experimental, or
standards-track at the discretion of the IESG. It MUST contain the standards-track at the discretion of the IESG. It MUST contain the
following information: following information:
(a) Name of the parameter type. (a) Name of the parameter type.
(b) Detailed description of the structure of the parameter field. This (b) Detailed description of the structure of the parameter field. This
skipping to change at page 86, line 4 skipping to change at page 91, line ?
structure MUST conform to the general type-length-value format structure MUST conform to the general type-length-value format
described in section 2.2.1. described in section 2.2.1.
(c) Detailed definition of each component of the parameter value. (c) Detailed definition of each component of the parameter value.
(d) Detailed description of the intended use of this parameter type, (d) Detailed description of the intended use of this parameter type,
and an indication of whether and under what circumstances and an indication of whether and under what circumstances
multiple instances of this parameter type may be found within the multiple instances of this parameter type may be found within the
same chunk. same chunk.
Additional parameter type codes may be allocated initially from the Additional parameter type codes may be allocated initially from the
range 0x0000 through 0xFFFD. If this space is exhausted, extension range 0x0000 through 0xFFFD. If this space is exhausted, extension
codes shall be allocated in the range 0x0000 through 0xFFFF. Where an codes shall be allocated in the range 0x0000 through 0xFFFF. Where an
extension code has been allocated, the format of the parameter must extension code has been allocated, the format of the parameter must
conform to the following structure: conform to the following structure:
Stewart, et al [Page 87]
First word (32 bits): First word (32 bits):
contains the parameter type code 0xFFFF and parameter length as contains the parameter type code 0xFFFF and parameter length as
described in section 2.2.1. described in section 2.2.1.
Second word: Second word:
first octet MUST be all 1's (0xFF). Next octet MUST be all 0's first octet MUST be all 1's (0xFF). Next octet MUST be all 0's
(0x00). Final two octets contain the allocated extension code (0x00). Final two octets contain the allocated extension code
value. value.
The Value portion of the parameter, if any, follows the second word. The Value portion of the parameter, if any, follows the second word.
skipping to change at page 86, line 32 skipping to change at page 91, line ?
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1| Length | |1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1| Length |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|1 1 1 1 1 1 1 1|0 0 0 0 0 0 0 0| Extension Type Code | |1 1 1 1 1 1 1 1|0 0 0 0 0 0 0 0| Extension Type Code |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
\ \ \ \
/ Value / / Value /
\ \ \ \
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
11.3 IETF-defined Additional Error Causes 12.3 IETF-defined Additional Error Causes
Additional cause codes may be allocated in the range 0x0004 to 0xFFFF Additional cause codes may be allocated in the range 0x0004 to 0xFFFF
upon receipt of any permanently-available public documentation upon receipt of any permanently-available public documentation
containing the following information: containing the following information:
(a) Name of the error condition. (a) Name of the error condition.
(b) Detailed description of the conditions under which an SCTP (b) Detailed description of the conditions under which an SCTP
endpoint should issue an Operation Error with this cause code. endpoint should issue an Operation Error with this cause code.
(c) Expected action by the SCTP endpoint which receives an Operation (c) Expected action by the SCTP endpoint which receives an Operation
Error chunk containing this cause code. Error chunk containing this cause code.
(d) Detailed description of the structure and content of data fields (d) Detailed description of the structure and content of data fields
which accompany this cause code. which accompany this cause code.
The initial word (32 bits) of a cause code parameter MUST conform to The initial word (32 bits) of a cause code parameter MUST conform to
the format shown in section 2.3.9, i.e.: the format shown in section 2.3.9, i.e.:
-- first two octets contain the cause code value -- first two octets contain the cause code value
-- last two octets contain length of the cause parameter. -- last two octets contain length of the cause parameter.
11.4 Payload Protocol Identifiers 12.4 Payload Protocol Identifiers
Except for value 0x00000000 which is reserved by SCTP to indicate the Except for value 0x00000000 which is reserved by SCTP to indicate the
absence of a payload protocol identifier in a DATA chunk, SCTP will absence of a payload protocol identifier in a DATA chunk, SCTP will
not be responsible for standardizing or verifying any payload protocol not be responsible for standardizing or verifying any payload protocol
identifiers; SCTP simply receives the identifier from the upper layer identifiers; SCTP simply receives the identifier from the upper layer
and carries it with the corresponding payload data. and carries it with the corresponding payload data.
The upper layer, i.e, the SCTP user, SHOULD standardize any specific The upper layer, i.e, the SCTP user, SHOULD standardize any specific
protocol identifier with IANA if it is so desired. The use of any protocol identifier with IANA if it is so desired. The use of any
specific payload protocol identifier is out of the scope of SCTP. specific payload protocol identifier is out of the scope of SCTP.
12. Suggested SCTP Protocol Parameter Values Stewart, et al [Page 88]
13. Suggested SCTP Protocol Parameter Values
The following protocol parameters are RECOMMENDED: The following protocol parameters are RECOMMENDED:
RTO.Initial - 3 seconds RTO.Initial - 3 seconds
RTO.Min - 1 second RTO.Min - 1 second
RTO.Max - 60 seconds
RTO.Alpha - 1/8 RTO.Alpha - 1/8
RTO.Beta - 1/4 RTO.Beta - 1/4
Valid.Cookie.Life - 5 seconds Valid.Cookie.Life - 5 seconds
Association.Max.Retrans - 10 attempts Association.Max.Retrans - 10 attempts
Path.Max.Retrans - 5 attempts (per destination address) Path.Max.Retrans - 5 attempts (per destination address)
Max.Init.Retransmits - 8 attempts Max.Init.Retransmits - 8 attempts
'retrans.count' - counter (per destination address) 'retrans.count' - counter (per destination address)
'receiver.buffer' - variable (per peer endpoint) 'receiver.buffer' - variable (per peer endpoint)
IMPLEMENTATION NOTE: The SCTP implementation SHOULD allow ULP to IMPLEMENTATION NOTE: The SCTP implementation may allow ULP to
customize some of these protocol parameters (see Section 9). customize some of these protocol parameters (see Section 9).
13. Acknowledgments 14. Acknowledgments
The authors wish to thank Mark Allman, Richard Band, Scott Bradner, The authors wish to thank Mark Allman, Richard Band, Scott Bradner,
Ram Dantu, R. Ezhirpavai, Sally Floyd, Matt Holdrege, Henry Houh, Ram Dantu, R. Ezhirpavai, Sally Floyd, Matt Holdrege, Henry Houh,
Christian Huetima, Gary Lehecka, Lyndon Ong, Kelvin Porter, Heinz Christian Huetima, Gary Lehecka, John Loughney, Daniel Luan, Lyndon
Prantner, Jarno Rajahalme, A. Sankar, Greg Sidebottom, Brian Wyld, and Ong, Kelvin Porter, Heinz Prantner, Jarno Rajahalme, Ivan Rodreguez,
many others for their invaluable comments. A. Sankar, Greg Sidebottom, Brian Wyld, and many others for their
invaluable comments.
14. Authors' Addresses 15. Authors' Addresses
Randall R. Stewart Tel: +1-847-632-7438 Randall R. Stewart Tel: +1-847-632-7438
Motorola, Inc. EMail: rstewar1@email.mot.com Motorola, Inc. EMail: rstewar1@email.mot.com
1501 W. Shure Drive, #2315 1501 W. Shure Drive, #2315
Arlington Heights, IL 60004 Arlington Heights, IL 60004
USA USA
Qiaobing Xie Tel: +1-847-632-3028 Qiaobing Xie Tel: +1-847-632-3028
Motorola, Inc. EMail: qxie1@email.mot.com Motorola, Inc. EMail: qxie1@email.mot.com
1501 W. Shure Drive, #2309 1501 W. Shure Drive, #2309
skipping to change at page 88, line 5 skipping to change at page 91, line ?
13615 Dulles Technology Drive 13615 Dulles Technology Drive
Herndon, VA. 20171 Herndon, VA. 20171
USA USA
Chip Sharp Tel: +1-919-472-3121 Chip Sharp Tel: +1-919-472-3121
Cisco Systems Inc. EMail:chsharp@cisco.com Cisco Systems Inc. EMail:chsharp@cisco.com
7025 Kit Creek Road 7025 Kit Creek Road
Research Triangle Park, NC 27709 Research Triangle Park, NC 27709
USA USA
Stewart, et al [Page 89]
Hanns Juergen Schwarzbauer Tel: +49-89-722-24236 Hanns Juergen Schwarzbauer Tel: +49-89-722-24236
SIEMENS AG SIEMENS AG
Hofmannstr. 51 Hofmannstr. 51
81359 Munich 81359 Munich
Germany Germany
EMail: HannsJuergen.Schwarzbauer@icn.siemens.de EMail: HannsJuergen.Schwarzbauer@icn.siemens.de
Tom Taylor Tel: +1-613-736-0961 Tom Taylor Tel: +1-613-736-0961
Nortel Networks Nortel Networks
1852 Lorraine Ave. 1852 Lorraine Ave.
Ottawa, Ontario Ottawa, Ontario
Canada K1H 6Z8 Canada K1H 6Z8
EMail:taylor@nortelnetworks.com EMail:taylor@nortelnetworks.com
Ian Rytina Tel: Ian Rytina Tel: +61-3-9301-6164
Ericsson Australia EMail:ian.rytina@ericsson.com Ericsson Australia EMail:ian.rytina@ericsson.com
37/360 Elizabeth Street 37/360 Elizabeth Street
Melbourne, Victoria 3000 Melbourne, Victoria 3000
Australia Australia
Malleswar Kalla Tel: +1-973-829-5212 Malleswar Kalla Tel: +1-973-829-5212
Telcordia Technologies Telcordia Technologies
MCC 1J211R MCC 1J211R
445 South Street 445 South Street
Morristown, NJ 07960 Morristown, NJ 07960
skipping to change at page 88, line 45 skipping to change at page 91, line ?
4531G Boelter Hall 4531G Boelter Hall
Los Angeles, CA 90095-1596 Los Angeles, CA 90095-1596
USA USA
Vern Paxson Tel: +1-510-642-4274 x 302 Vern Paxson Tel: +1-510-642-4274 x 302
ACIRI EMail: vern@aciri.org ACIRI EMail: vern@aciri.org
1947 Center St., Suite 600, 1947 Center St., Suite 600,
Berkeley, CA 94704-1198 Berkeley, CA 94704-1198
USA USA
15. References 16. References
[1] Eastlake , D. (ed.), "Randomness Recommendations for Security", [1] Eastlake , D. (ed.), "Randomness Recommendations for Security",
RFC 1750, December 1994. RFC 1750, December 1994.
[2] Deutsch, P., and Gailly, J-L., "ZLIB Compressed Data Format [2] Deutsch, P., and Gailly, J-L., "ZLIB Compressed Data Format
Specification version 3.3", RFC 1950, May 1996. Specification version 3.3", RFC 1950, May 1996.
[3] Allman, M., Paxson, V., and Stevens, W., "TCP Congestion [3] Allman, M., Paxson, V., and Stevens, W., "TCP Congestion
Control", RFC 2581, April 1999. Control", RFC 2581, April 1999.
Stewart, et al [Page 90]
[4] Rivest, R., "The MD5 Message-Digest Algorithm", RFC 1321, [4] Rivest, R., "The MD5 Message-Digest Algorithm", RFC 1321,
August 1999. August 1999.
[5] Allman, M., and Paxson, V., "On Estimating End-to-End Network [5] Allman, M., and Paxson, V., "On Estimating End-to-End Network
Path Properties", Proc. SIGCOMM'99, 1999. Path Properties", Proc. SIGCOMM'99, 1999.
[6] Karn, P., and Simpson, W., "Photuris: Session-Key Management [6] Karn, P., and Simpson, W., "Photuris: Session-Key Management
Protocol", RFC 2522, March 1999. Protocol", RFC 2522, March 1999.
[7] Bradner, S., "The Internet Standards Process -- Revision 3", [7] Bradner, S., "The Internet Standards Process -- Revision 3",
skipping to change at page 89, line 41 skipping to change at page 91, line ?
[13] Fraser, B. (ed.), "Site Security Handbook", RFC 2196, September [13] Fraser, B. (ed.), "Site Security Handbook", RFC 2196, September
1997. 1997.
[14] Kent, S., and Atkinson, R., "Security Architecture for the [14] Kent, S., and Atkinson, R., "Security Architecture for the
Internet Protocol", RFC 2401, November 1998. Internet Protocol", RFC 2401, November 1998.
[15] Savage, S., Cardwell, N., Wetherall, D., and Anderson, T., [15] Savage, S., Cardwell, N., Wetherall, D., and Anderson, T.,
"TCP Congestion Control with a Misbehaving Receiver", ACM "TCP Congestion Control with a Misbehaving Receiver", ACM
Computer Communication Review, 29(5), October 1999. Computer Communication Review, 29(5), October 1999.
This Internet Draft expires in 6 months from January, 2000 This Internet Draft expires in 6 months from February, 2000
 End of changes. 

This html diff was produced by rfcdiff 1.23, available from http://www.levkowetz.com/ietf/tools/rfcdiff/