home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Internet Info 1993
/
Internet Info CD-ROM (Walnut Creek) (1993).iso
/
inet
/
internet-drafts
/
draft-ietf-pppext-cipx-02.txt
< prev
next >
Wrap
Text File
|
1993-04-13
|
46KB
|
1,433 lines
Internet Area - PPPEXT Working Group Saroop Mathur
Internet Draft Mark S. Lewis
Telebit Corp.
Expires in six months April 1993
Compressing IPX Headers Over WAN Media (CIPX)
Status of this Memo
This draft is a proposed standard protocol for the Internet
community and requests discussion and suggestions for improvement.
This document is being submitted to the Internet Engineering Task
Force (IETF) through the Point-to-Point Protocol Extensions
(PPPEXT) Working Group. Comments should be sent to the authors and
the ietf-ppp@ucdavis.edu mailing list.
Distribution of this memo is unlimited.
This document is an Internet Draft. Internet Drafts are working
documents of the IETF, its Areas, and its Working Groups. Internet
Drafts are valid for a maximum of six months. Internet Drafts may
be updated, replaced, or obsoleted by other documents at any time.
Abstract
This document describes a method for compressing the headers of IPX
datagrams (CIPX). With this method, it is possible to
significantly improve performance over lower speed wide area
network (WAN) media. For normal IPX packet traffic, CIPX can
provide a compression ratio of approximately 2:1 including both IPX
header and data. This method can be used on various type of WAN
media, including those supporting PPP and X.25.
Mathur, Lewis expires in six months [Page 1]
DRAFT CIPX April 1993
Introduction
Internetwork Packet Exchange (IPX) is a protocol defined by the
Novell Corporation. It is derived from the Internet Datagram
Protocol (IDP) protocol of the Xerox Network Systems (XNS) family
of protocols. IPX is a datagram, connectionless protocol that does
not require an acknowledgment for each packet sent. The IPX
protocol corresponds to the network layer of the ISO model.
Usually, there is a transport layer protocol above IPX. The most
common transport protocol is the Netware Core Protocol (NCP), which
is used for file server access. The Sequenced Packet Exchange
(SPX) is the reliable connection-based transport protocol commonly
used by applications.
The IPX packet consists of a 30 octet IPX header, usually followed
by the transport layer protocol header. The NCP header is 6 octets
in length. The SPX header is 12 octets in length.
Two strategies are described below for compressing IPX headers.
This specification requires that implementations of CIPX support
both IPX header compression strategies. These header compression
algorithms are based on those Van Jacobson described [1] for TCP/IP
packets.
The first strategy is to compress only the IPX header. This
compression algorithm can be used to compress any IPX packet,
without affecting the transport protocol. This algorithm
compresses a 30 octet IPX header into a one to seven octet header.
The second strategy is to compress the combined IPX and NCP
headers. This algorithm compresses only NCP packets with NCP type
of 0x2222 and 0x3333. This algorithm compresses a 36 octet NCP/IPX
header into a one to eight octet header.
Lastly, note that it is possible, and many times desirable, to use
this type of header compression in conjunction with some type of
data compression. After intelligently compressing the packet
header, data compression can be effective in reducing packet size
further. It is important that data compression should be done
after header compression. Conversely, data decompression should be
done before header decompression.
Mathur, Lewis expires in six months [Page 2]
DRAFT CIPX April 1993
IPX Compression Algorithm
The normal IPX header consists of the following fields: checksum,
packet length, transport control (hop count), packet type,
destination and source address fields.
+-----------------------+
| Checksum |
+-----------------------+
| Packet Length |
+-----------+-----------+
| Hops |Packet Type|
+-----------+-----------+
| Destination |
| Address |
| (12 Octets) |
+-----------------------+
| Source |
| Address |
| (12 Octets) |
+-----------------------+
IPX PACKET HEADER
The IPX header diagram above is shown without the unnecessary field
alignment details. Consider each field of the IPX header
separately, and how it typically changes.
Historically, Novell has not used the Checksum field in the IPX
header, and has required that this field be set to 0xFFFF. Since
the Checksum field remains constant, it is clear that the value can
be compressed.
Where Checksums are implemented (not 0xFFFF), the Checksum MUST be
included in the compressed packet. Recalculating the checksum
would destroy the end-to-end reliability of the connection. Note
that Checksums are now implemented in the Fault Tolerant Servers.
For most links, the Packet Length can be determined from the MAC
layer. There are cases in which the length cannot be determined
from the MAC layer. For example, some hardware devices pad packets
to a required minimum length. For links where it is not possible
to determine the IPX packet length from the MAC layer, packet
length needs to be included in the compressed packet.
The Transport Control (Hops) field usually does not change between
Mathur, Lewis expires in six months [Page 3]
DRAFT CIPX April 1993
two end-points. For the purposes of compression, we will assume
that it never changes, and will not examine this field when
determining a connection.
The Packet Type field is constant for any connection.
The Destination and Source Address fields are each made up of 12
octets: Network (4 octets), Node (6 octets), and Socket (2 octets)
fields. For any specific IPX connection, the Destination and
Source Address fields are constant.
Hence, the fields that may need to be included in the compressed
IPX header are the Checksum and the Packet Length.
While using this IPX header compression algorithm, packets can be
lost. The loss of an Initial packet presents a problem. In this
case, if the sender later tries to send a compressed packet, the
receiving end cannot decompress the packet correctly.
Sufficient information is not available in the IPX header to
determine when a re-transmission has occured. For this reason, it
is necessary that the sender of an Initial packet be guaranteed
that the packet has been received. Therefore, we provide a
mechanism for Confirmation of an Initial packet.
Mathur, Lewis expires in six months [Page 4]
DRAFT CIPX April 1993
NCP/IPX Header Compression
Since most IPX packets are Netware Core Protocol packets (packet
type 17), compressing the NCP header will give us added
performance. A minimal CIPX implementation MUST also implement
NCP/IPX compression.
+------------+
| NCP |
| Type |
+------------+
| Sequence |
| Number |
+------------+
| Connection |
|(low octet) |
+------------+
| Task |
| Number |
+------------+
| Connection |
|(high octet)|
+------------+
NCP HEADER
The NCP header is 6 octets in length consisting of the following
fields: NCP type, sequence number, connection number and task
number.
The NCP type field values that are currently defined are:
1111 Create Connection
2222 NCP request from workstation
3333 NCP reply from file server
5555 Destroy Connection
7777 Burst Mode Packet
9999 Server Busy Packet
This NCP header compression algorithm only compresses packets that
have a type field value of 0x2222 or 0x3333. All other types of
packets are not compressed at the NCP level.
If NCP type is 0x2222, then this packet is a request from the
Mathur, Lewis expires in six months [Page 5]
DRAFT CIPX April 1993
client to the server. Conversely if NCP type is 0x3333, this is a
response from the server to the client.
The connection number is a constant for a given connection.
The sequence number is increased by one for each new request.
Hence the sequence number can be determined implicitly. The
decompressor increments the sequence number for each compressed
packet it receives for a connection.
The task number can change unpredictably, although it might remain
constant for several packets. If the NCP task number is different
from the last one for this connection, the NCP task number must be
included.
If the NCP packet is lost, it will be retransmitted through the
normal transport layer mechanisms. The Initial NCP packet does not
require confirmation, as a re-transmitted packet can be easily
identified. This is accomplished by comparing the sequence number
of the packet to the sequence number of the previous packet. If
the sequence number is not exactly one greater than the previous
packet, a new Initial packet must be sent, although the same
connection slot may be used.
In the event of compressed packet loss, the sequence number will be
too small. When the IPX Checksum is present, the loss can be
determined at the destination system by an incorrect checksum.
When there is no checksum present, the loss is more likely to be
detected upon receiving a later retransmission.
NCP Burst Mode Packets
The burst mode protocol uses the NCP type value of 0x7777. This
type of packet does not have the normal NCP header described above.
Instead, it has a 36 octet burst header. The above NCP header
compression algorithm should not be used to compress this packet.
The IPX header in this packet is still compressible with the IPX
header compression algorithm described.
SPX Packets
SPX packets are typically used by applications which require
reliable service such as print servers. It is possible to apply a
similar NCP/IPX technique to SPX/IPX packets. At this time, we
have not described such a mechanism. The IPX header in this packet
is still compressible with the IPX header compression algorithm
described.
Mathur, Lewis expires in six months [Page 6]
DRAFT CIPX April 1993
Compression Header
IPX compression should be negotiated by some means (eg. IPXCP or
IPXWAN). Each end must negotiate the desired options, such as the
maximum number of concurrent connections which will be maintained
in each direction. Once IPX compression is negotiated, all IPX
packets sent over that link have a CIPX header added to the
beginning of the packet.
The one octet CIPX header is added even when a regular IPX packet
is sent over the link. By including the CIPX header on every
packet, we support the ability to run CIPX over various WAN links
as if it were a normal IPX packet. It does not rely on any new
link specific packet demultiplexing.
Implementations of this compression protocol must maintain send and
receive tables indicating the state of each connection. The
original header for each connection is stored in a "slot".
Typically, each client-server connection will use a separate slot.
Both sides keep a copy of the full IPX header corresponding to each
slot. The sending side (compressor) uses this information to
determine the fields that have changed. The receiving side
(decompressor) uses this information to reconstruct the original
packet header.
The CIPX packet header specifies the type of the packet and any
options for that packet. The basic header is one octet in length.
7 6 5 4 3 2 1 0
+---+---+---+---+---+---+---+---+
| | | | | | | | |
+---+---+---+---+---+---+---+---+
^ ^ ^ ^ ^ ^ ^ ^
| | | | | | | |
| | | | |___|___|___|___ Packet Type
| | | | 0 Compressed
| | | | 1 Regular
| | | | 3 Confirmed Initial
| | | | 5 Confirm
| | | | 7 Unconfirmed Initial
| | | | 9 Reject
| | | | 11-15 Reserved
| | | |
|__ |__ |__ |___________________ Packet Type Dependent Flags
FLAGS OCTET
Mathur, Lewis expires in six months [Page 7]
DRAFT CIPX April 1993
As can be seen above, the low order bits specify the packet type.
All Compressed packets have a lowest bit of zero. The other packet
types are odd numbers.
Note that the Flags octet MUST NOT contain the value 0xFF. This is
necessary to distinguish the CIPX flags octet from a normal IPX
header with a 0xFFFF checksum field. It is important to be able to
recognize a normal IPX header regardless of the state of
compression. It is possible with some link layer protocols such as
X.25 Permanent Virtual Circuits that one end of the link may fail
and start sending IPX packets.
Each packet type has associated flag bits, which are called Packet
Type Dependent Flags. Different packet types have different Packet
Type Dependent Flags. All bits that are reserved or are not
specified must be set to zero.
Since none of the packet types other than Compressed currently uses
any of the flag bits, the packet type field could easily be
expanded. Any future expansion must ensure that at least one of
the bits remains zero.
Mathur, Lewis expires in six months [Page 8]
DRAFT CIPX April 1993
Compressed Packet
This type of packet does not contain a packet header (either 30
byte IPX, or 36 byte NCP). A slot number indicates to the receiver
which saved header to use to formulate the original packet header
before passing the packet up to IPX.
7 6 5 4 3 2 1 0
+---+---+---+---+---+---+---+---+
| | | | | 0 | 0 | 0 | 0 |
+---+---+---+---+---+---+---+---+
^ ^ ^ ^ ^ ^ ^ ^
| | | | | | | |
| | | | | | | |___ Packet Type
| | | | | | | 0 Compressed Packet
| | | | | | |
| | | | |___|___|_______ Reserved (Must be zero)
| | | |
| | | |___________________ Task Number (NCP only)
| | | 0 Assume same as last packet
| | | 1 Included in packet
| | |
| | |_______________________ Length
| | 0 Determine from MAC length
| | 1 Included in packet
| |
| |___________________________ Checksum
| 0 Assume 0xFFFF
| 1 Included in packet
|
|_______________________________ Slot Number
0 Assume same as last packet
1 Included in packet
Consider each flag in the flags octet, looking at the high order
bits working toward the lower order bits. Each of the fields is
optional, but if present will be found in the same order in the
compressed packet header.
Slot Number
The slot number flag indicates the slot number field is
included in the compressed packet. The slot number field is
one octet in length and specifies the number of the slot which
corresponds to the Initial packet header. Slots are numbered
starting at zero and continue to the maximum number of slots
Mathur, Lewis expires in six months [Page 9]
DRAFT CIPX April 1993
minus one.
By default, slot compression is disabled, and every packet has
the slot number included. If negotiated, slot compression can
be enabled for those slots which were created by the
Unconfirmed Initial packet. When slot compression is enabled
and the packet will be compressed using the same slot as for
the last packet, the compressor clears the slot number flag
and omits the slot number field.
Implementation Note:
Slot compression MUST only be enabled in a receiver which
can account for all erroneous and discarded packets.
When a packet has been discarded, the slot number is
indeterminate for future packets. The decompressor MUST
discard all further packets until a slot number is
received.
Checksum
If the checksum flag is set, the compressed packet will
include the 2 octet checksum. If this flag is not set, then
the decompressor assumes the checksum is 0xFFFF. Note that
meaningful checksums must be included in the packet with the
flag set.
Length
The length flag indicates the inclusion of the IPX data length
field in the compressed packet. This is the Length field from
the original IPX packet header. It specifies the length of
IPX header and data in the packet prior to compression. It
does not include the CIPX compression field such as flags,
slot number, checksum, length field, or the NCP task number.
Note that it is preferable to determine the length field from
the MAC layer whenever possible, by subtracting the length of
the compression header fields and adding the length of the
saved packet header.
Since every octet is significant over lower speed WAN links,
an optimization is used in the specification of the length.
It can be specified as a one, two or three octet field. If
the length is in the range 0 to 127, then it is specified as a
one octet field. If the length is in the range 128 to 16383,
it is specified as a two octet field in high to low order,
with the first octet in the range 128 to 191. Otherwise, if
the length is greater than 16383, the first octet contains
Mathur, Lewis expires in six months [Page 10]
DRAFT CIPX April 1993
192, and the second and third octets contain the full length.
(This scheme is extensible to 8 octets, but currently is not
required in the IPX protocol suite.)
+-+-+-+-+-+-+-+-+
|0| length | length < 128
+-+-+-+-+-+-+-+-+
ONE OCTET LENGTH FIELD
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|1 0| length | length < 16384
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
TWO OCTET LENGTH FIELD
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
|1 1 0 0 0 0 0 0| length | length < 65535
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
THREE OCTET LENGTH FIELD
Task Number
The NCP task number flag indicates the NCP task number is
included in the compressed packet (see NCP/IPX compression
above). Otherwise, we use the same NCP task number as that of
last packet.
Based upon the bits set in the flags octet, optional portions are
included in the compressed IPX header. The minimum compressed IPX
header contains only the Flags octet. All fields in the original
IPX header have been compressed out of the header. The maximum
compressed IPX header can include up to 7 octets, the Flags, Slot,
Checksum (2 octets), and Length (3 octets) fields, or 8 octets if
the NCP Task Number is included. The minimum and maximum
compressed IPX packets are shown below. Header fields are one
octet in length except where noted.
Mathur, Lewis expires in six months [Page 11]
DRAFT CIPX April 1993
+--------+---------
| Flags | DATA ...
| 0x00 |
+--------+---------
MINIMUM COMPRESSED IPX PACKET
+--------+--------+---------+---------+---------
| Flags | Slot |Checksum | Length | DATA ...
| 0xE0 | Number |2 octets |3 octets |
+--------+--------+---------+---------+---------
MAXIMUM COMPRESSED IPX PACKET
+--------+--------+---------+---------+--------+---------
| Flags | Slot |Checksum | Length |NCP Task| DATA ...
| 0xF0 | Number |2 octets |3 octets | Number |
+--------+--------+---------+---------+--------+---------
MAXIMUM COMPRESSED NCP/IPX PACKET
Mathur, Lewis expires in six months [Page 12]
DRAFT CIPX April 1993
Regular Packet
The Regular packet type designates an IPX packet for which no
compression is desired. This type of packet is sent when a packet
cannot be compressed, or a decision is made not to compress it.
7 6 5 4 3 2 1 0
+---+---+---+---+---+---+---+---+
| 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 |
+---+---+---+---+---+---+---+---+
^ ^ ^ ^ ^ ^ ^ ^
| | | | | | | |
| | | | |___|___|___|___ Packet Type
| | | | 1 Regular
| | | |
|__ |__ |__ |___________________ Reserved (must be zero)
The Regular packet is rarely sent. Usually, the Regular packet is
sent when there is not enough memory for the overhead of a new
compression slot. Also, this type is included for future
unforeseen changes to the IPX protocol which defeat the
effectiveness of compression.
Implementation Note:
The Regular Packet can be used for packets that are sporadic,
which are not worth setting-up a compression slot. This may
be hard to determine for specific protocols. Various methods
such as hold-down and least-recently-used timers are currently
being used.
On receipt, the 1 octet header is simply removed and the packet
passed up to IPX.
The entire IPX packet follows the single Flags octet. Note for a
Regular Packet (not compressed or uncompressed), the slot number
field is not included.
Mathur, Lewis expires in six months [Page 13]
DRAFT CIPX April 1993
Confirmed Initial Packet
The Confirmed Initial packet type is used by the compressor to
inform the decompressor of the original packet header which will be
used for subsequent compression, and to request Confirmation. The
high order 4 bits are reserved for expansion to support additional
protocols.
7 6 5 4 3 2 1 0
+---+---+---+---+---+---+---+---+
| 0 | 0 | 0 | 0 | 0 | 0 | 1 | 1 |
+---+---+---+---+---+---+---+---+
^ ^ ^ ^ ^ ^ ^ ^
| | | | | | | |
| | | | |___|___|___|___ Packet Type
| | | | 3 Confirmed Initial
| | | |
|__ |__ |__ |___________________ 0 IPX Protocol
1-15 Reserved
This type of packet is sent to inform the receiver to associate the
IPX packet header with a slot number. This packet is sent each
time a different header format is sent for a given slot, or when
the sender has not received a Confirmation Packet from the
receiver.
The Flags octet lower 4 bits indicate the Confirmed Initial CIPX
packet type. The high order 4 bits are reserved for expansion to
support additional protocols. The Flags octet is always followed
by the Slot Number and an ID field.
+---------+---------+---------+------------+----------
| Flags | Slot | ID | IPX | DATA ...
| 0x03 | Number | | Header |
+---------+---------+---------+------------+----------
CONFIRMED INITIAL PACKET
For each slot, the ID will increment with every new header sent.
Different slots may have the same ID. The combination of slot and
ID uniquely identify a header. In practice, the ID octet can be
any number which is unique for a "reasonably long period" of time.
A reasonably long period is a function of transmission speed, round
trip delays, and network load. There must be very little chance of
Mathur, Lewis expires in six months [Page 14]
DRAFT CIPX April 1993
duplicate slot and ID combinations within this period. Otherwise,
there is ambiguity as to which header is being identified.
Implementation Note:
With the use of the ID octet, an implementation does not have
to hold a packet forever. If a Confirmed Initial packet goes
unconfirmed for too long a period, the same packet may be sent
using another slot. This allows an implementation to continue
through the datastream without having to retransmit the same
Confirmed Initial packet using the same slot.
For many implementations, it may be better to resend the same
Confirmed Initial packet until a corresponding Confirm packet
is received. Using this method, an implementation may avoid
some of the complexity of choosing which packet to send using
a slot for which a confirmation has not been received within
the given period. The implementation simply resends the same
packet with the same ID until confirmed. The choice between
these two methods is left as an implementation detail.
Note that a Confirmed Initial header is followed by a complete IPX
packet.
Mathur, Lewis expires in six months [Page 15]
DRAFT CIPX April 1993
Confirm Packet
The Confirm packet type is used by the decompressor to tell the
compressor that it has received the Confirmed Initial packet.
When the compressor receives this, it can start sending Compressed
frames.
7 6 5 4 3 2 1 0
+---+---+---+---+---+---+---+---+
| 0 | 0 | 0 | 0 | 0 | 1 | 0 | 1 |
+---+---+---+---+---+---+---+---+
^ ^ ^ ^ ^ ^ ^ ^
| | | | | | | |
| | | | |___|___|___|___ Packet Type
| | | | 5 Confirm
| | | |
|__ |__ |__ |___________________ Reserved (must be zero)
A Confirm Packet is exactly 3 octets in length. It consists of the
Flags, Slot Number and ID fields. The Slot Number field contains
the number of the slot which is being acknowledged. The ID field
contains the ID of the Confirm Initial Packet which is being
acknowledged.
+---------+---------+----------+
| Flags | Slot | ID |
| 0x05 | Number | |
+---------+---------+----------+
CONFIRM PACKET
Mathur, Lewis expires in six months [Page 16]
DRAFT CIPX April 1993
Unconfirmed Initial Packet
The Unconfirmed Initial packet type is used by the compressor to
inform the decompressor of the original packet header which will be
used for subsequent compression while not requesting confirmation.
After sending an Unconfirmed Initial packet, the compressor may
immediately send Compressed packets without confirmation.
7 6 5 4 3 2 1 0
+---+---+---+---+---+---+---+---+
| 0 | 0 | 0 | 0 | 0 | 1 | 1 | 1 |
+---+---+---+---+---+---+---+---+
^ ^ ^ ^ ^ ^ ^ ^
| | | | | | | |
| | | | |___|___|___|___ Packet Type
| | | | 7 Unconfirmed Initial
| | | |
|__ |__ |__ |___________________ 0 NCP Protocol
1-15 Reserved
This type of packet is sent to inform the receiver to associate the
IPX packet header with a slot number. This packet is sent each
time a different header format is sent for a given slot.
The Flags octet lower 4 bits indicate the Unconfirmed Initial CIPX
packet type. The high order 4 bits are reserved for expansion to
support additional protocols. The Flags octet is always followed
by the Slot Number.
+---------+---------+------------+-----------+---------
| Flags | Slot | IPX | NCP | NCP
| 0x07 | Number | Header | Header | DATA ...
+---------+---------+------------+-----------+---------
UNCONFIRMED INITIAL PACKET
Note that an Unconfirmed Initial header is followed by a complete
IPX packet.
Mathur, Lewis expires in six months [Page 17]
DRAFT CIPX April 1993
Reject Packet
The Reject packet type is used by the decompressor to tell the
compressor that it has received an Initial packet (Confirmed or
Unconfirmed) with a Flag value which it does not support. This is
provided for future expansion. Any CIPX packet type which is
unknown to the decompressor MUST be rejected.
When the compressor receives this, it should record the IPX Packet
Type as uncompressible. Future packets of this type should be sent
with only IPX header compression.
7 6 5 4 3 2 1 0
+---+---+---+---+---+---+---+---+
| 0 | 0 | 0 | 0 | 1 | 0 | 0 | 1 |
+---+---+---+---+---+---+---+---+
^ ^ ^ ^ ^ ^ ^ ^
| | | | | | | |
| | | | |___|___|___|___ Packet Type
| | | | 9 Reject
| | | |
|__ |__ |__ |___________________ Reserved (must be zero)
A Reject Packet is exactly 3 octets in length. It consist of the
Flags, Slot Number and Rejected Flags fields. The Slot Number
field contains the number of the slot of the packet which is being
rejected.
The Rejected Flags field contains the flag(s) of the unsupported
packet type and/or packet dependent option. This octet will contain
the 4-bit packet type of the packet being rejected. It may also
include a specific combination of packet type dependent flag(s)
that is being rejected. The intention of this field is to provide
information to the sender about why the packet is being rejected.
Intelligent implementations can use this information to adjust
their use of certain packet types or options.
Implementation Note:
The Flags value of 0xFF is not a valid CIPX packet type.
Hence, such a packet type should be recognized as a standard
IPX header and forwarded without CIPX processing to the
appropriate routines. Under no circumstances should a Flags
value of 0xFF be rejected in a Reject Packet.
Mathur, Lewis expires in six months [Page 18]
DRAFT CIPX April 1993
+---------+---------+----------+
| Flags | Slot | Rejected |
| 0x09 | Number | Flags |
+---------+---------+----------+
REJECT PACKET
Mathur, Lewis expires in six months [Page 19]
DRAFT CIPX April 1993
Compression Negotiation over PPP Links
For PPP links [2], the use of header compression can be negotiated
by IPXCP [3]. By default, no compression is enabled.
The IPX-Compression-Protocol Configuration Option is used to
indicate the ability to receive compressed packets. Each end of
the link must separately request this option if bi-directional
compression is desired.
The PPP Protocol field is set to the same value as the usual IPX
packets, and all IPX packets sent over the link MUST conform to the
compressed format.
A summary of the IPX-Compression-Protocol Configuration Option
format to negotiate Telebit IPX header compression (CIPX) is shown
below. The fields are transmitted from left to right.
0 1 2 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Type | Length | IPX-Compression-Protocol |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Max-Slot-Id | Options |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Type
3
Length
6
IPX-Compression-Protocol
0002 (hex) for Telebit Compressed IPX headers (CIPX).
Max-Slot-Id
The Max-Slot-Id field is one octet and indicates the maximum
slot identifier. This is one less than the actual number of
slots; the slot identifier has values from zero to Max-Slot-
Id.
Mathur, Lewis expires in six months [Page 20]
DRAFT CIPX April 1993
Options
The Options field is one octet, and is comprised of the
"logical or" of the following values:
0 No options.
1 The slot identifer may be compressed.
The slot identifier must not be compressed if there is no
ability for the PPP link level to indicate an error in
reception to the decompression module. Synchronization
after errors depends on receiving a packet with the slot
identifier.
2 Redefine Compressed Packet type bits 1-3.
It was noted earlier that packet types have been chosen
such that only the Compressed Packet type is an even
number value with the lowest order bit of zero. All
other packet types are odd values with a lowest order bit
of one. The reason for this assignment was to make it
possible to determine the Compressed Packet type by
examining only one bit. This make it possible to use all
the other 7 bits to indicate status in the Compressed
Packet. The 7 bits are composed of the upper 4 bits
which are permanently defined to indicate packet
dependent flags, plus bits 1-3 which are otherwise part
of the Packet Type. The upper 4 bits are defined above.
The redefinition of bits 1-3 of the Compressed Packet
type is left for future expansion.
7 6 5 4 3 2 1 0
+---+---+---+---+---+---+---+---+
| | | | | | | | 0 |
+---+---+---+---+---+---+---+---+
^ ^ ^ ^ ^ ^ ^ ^
| | | | | | | |___ Packet Type
| | | | | | | 0 Compressed Packet
| | | | | | |
| | | | |___|___|_______ Redefined bits
| | | |
|___|___|___|___________________ Compressed Packet flags
By default, this feature in not enabled and this flag is
set to zero. When this flag is set to one, it indicates
the desire to use this feature.
Mathur, Lewis expires in six months [Page 21]
DRAFT CIPX April 1993
Compression Negotiation over IPXWAN Links
"IPXWAN" is the protocol Novell uses to exchange necessary router
to router information prior to exchanging standard IPX routing
information and traffic over WAN datalinks [4]. To negotiate the
Telebit compression option, we add an option to the IPXWAN timer
request/response packet.
The Timer Request packet contains the following Telebit compression
option:
WOption Number 80 - Define compression type
WAccept Option 01 - 0=No, 1=Yes, 3=N/A
WOption Data Len 00 03 - Length of option
WOption Data 00 - Telebit's compression (CIPX)
WOption Data XX - Compression options
WOption Data NN - Compression slots
Where the WOption Data fields are:
00 Telebit's compression option described in this
document (CIPX).
XX Compression options as defined below:
0x01 Compress slot ID when possible
0x02 Redefine Compressed Packet type bits 1-3.
NN The requested # of compression slots.
Accept Option (for compression type) must be set to YES if the
option is supported and NO if the option is not supported. A Timer
Response must respond with only one header compression type set to
YES.
The Timer Response packet that accepts the option will look like
this:
WOption Number 80 - Define compression type
WAccept Option 01 - 0=No, 1=Yes, 3=N/A
WOption Data Len 00 03 - Length of option
WOption Data 00 - Telebit's compression (CIPX)
WOption Data XX - Compression options
WOption Data NN - Compression slots
Mathur, Lewis expires in six months [Page 22]
DRAFT CIPX April 1993
Where the WOption Data fields are:
00 Telebit's compression option described in this
document (CIPX).
XX Compression options as defined below:
0x01 Compress slot ID when possible
0x02 Redefine Compressed Packet type bits 1-3.
NN The negotiated # of slots (The lower of each side's
requested number of slots)
IPX packets (except of course IPXWAN packets) are not sent over the
link until the IPXWAN negotiations are completed. Once IPXWAN
negotiations are completed, regular IPX packets can be sent over
the link.
If both ends of the link agree on the compression options, then the
IPX packets are sent using the specified options. If either end of
the link does not accept a compression option, then this
compression option will not be used. Compression will be done
using any remaining options. Options, by definition, are not
required. Implementations MUST support CIPX without any options.
It is the responsibility of the router sending the IPXWAN Timer
Response to inform the other router of the options that will be
used. The Timer Response MUST contain a subset of the options
received in a Timer Request.
To be clear, IPXWAN is used to set up a symmetrical compression
link. Compression is configured identically in both directions.
Each end will use the same number of slots and same compression
options. It is illegal for link ends to use different number of
slots or different options.
Mathur, Lewis expires in six months [Page 23]
DRAFT CIPX April 1993
IPX Compression Performance
The performance of this algorithm will depend on the number of
active connections and the number of slots negotiated. If the
number of slots is greater than the number of connections, the hit
rate should be very high giving a very high compression ratio. The
performance also depends on the average size of the IPX packets.
If the average size of packets is small, then compression will
result in a more noticeable performance improvement.
avg_data_len + uncomp_header_len
Compression ratio = ----------------------------------
avg_data_len + avg_comp_header_len
Where 'avg_data_len' is the average length of data in the IPX
packet, and 'uncomp_head_len' is the uncompressed header length
which is fixed at 30 octets. Where 'avg_comp_header_len' is the
average length of the compressed IPX header. The length of the
minimum compressed IPX header is 1 octet. The length of the
maximum compressed NCP/IPX header is 8 octets (including the NCP
task number), but since nothing yet sends packets with a length
greater than 16K, 7 octets is the reality. Perhaps a reasonable
'avg_comp_header_len' is 2, assuming the inclusion of the flag and
slot number octets.
The maximum length of the data in an IPX packet is 546 octets (576
octets - 30 octet IPX header), although newer implementations may
send packets of up to 4096 octets. The minimum length of the data
in an IPX packet is 1 octet. Within the normal distribution of
small NCP packets, perhaps a reasonable 'avg_data_len' is 26
octets.
Mathur, Lewis expires in six months [Page 24]
DRAFT CIPX April 1993
546 + 30
Minimal Compression = -------- = 1.04
546 + 6
1 + 30
Maximal Compression = ------ = 15.50
1 + 1
26 + 30
Likely Compression = ------- = 2.00
26 + 2
Security Considerations
Security issues are not discussed in this memo.
Mathur, Lewis expires in six months [Page 25]
DRAFT CIPX April 1993
References
1 Jacobson, Van, "Compressing TCP/IP Headers for Low-Speed Serial
Links", RFC 1144, February 1990
2 Simpson, W. A., "The Point-to-Point Protocol (PPP)", RFC 1331, May
1992.
3 Simpson, W. A., "The PPP Internet Packet Exchange Control Protocol
(IPXCP)", an internet draft which is work in process, December 1992
4 Allen, Michael, "Novell IPX Over Various WAN Media [IPXWAN]", RFC
1362, August 1992
Acknowledgements
This compression algorithm incorporates many ideas from the Van
Jacobson TCP/IP header compression algorithm.
Michael Allen from Novell provided a lot of valuable feedback in
the design of this algorithm. Marty Del Vecchio at Shiva Corp.
made a couple of good observations. Bill Simpson was, as always,
very helpful.
Authors' Address
Saroop Mathur
Telebit Corp.
1315 Chesapeake Terrace
Sunnyvale, CA 94089-1100
email: mathur@telebit.com
Mark S. Lewis
Telebit Corp.
1315 Chesapeake Terrace
Sunnyvale, CA 94089-1100
email: Mark.S.Lewis@telebit.com