ESP CBC Ciphers: The ESP CBC-Mode Cipher Algorithms
One common recipe for using block ciphers such as Blowfish, CAST-128, IDEA, RC5 and 3DES with IPsec ESP.
Pereira, R. and R. Adams, "The ESP CBC-Mode Cipher Algorithms", RFC 2451, DOI 10.17487/RFC2451, November 1998, <https://www.rfc-editor.org/info/rfc2451>.
@misc{rfc2451,
series = {Request for Comments},
number = 2451,
howpublished = {RFC 2451},
publisher = {RFC Editor},
doi = {10.17487/RFC2451},
url = {https://www.rfc-editor.org/info/rfc2451},
author = {R. Pereira and R. Adams},
title = {{The ESP CBC-Mode Cipher Algorithms}},
pagetotal = 14,
year = 1998,
month = nov,
} RFC text
Contents
Network Working Group R. Pereira
Request for Comments: 2451 TimeStep Corporation
Category: Standards Track R. Adams
Cisco Systems Inc.
November 1998
The ESP CBC-Mode Cipher Algorithms
Status of this Memo
This document specifies an Internet standards track protocol for the
Internet community, and requests discussion and suggestions for
improvements. Please refer to the current edition of the "Internet
Official Protocol Standards" (STD 1) for the standardization state
and status of this protocol. Distribution of this memo is unlimited.
Copyright Notice
Copyright (C) The Internet Society (1998). All Rights Reserved.
Abstract
This document describes how to use CBC-mode cipher algorithms with
the IPSec ESP (Encapsulating Security Payload) Protocol. It not only
clearly states how to use certain cipher algorithms, but also how to
use all CBC-mode cipher algorithms.
Table of Contents
1. Introduction...................................................2
1.1 Specification of Requirements...............................2
1.2 Intellectual Property Rights Statement......................2
2. Cipher Algorithms..............................................2
2.1 Mode........................................................3
2.2 Key Size....................................................3
2.3 Weak Keys...................................................4
2.4 Block Size and Padding......................................5
2.5 Rounds......................................................6
2.6 Backgrounds.................................................6
2.7 Performance.................................................8
3. ESP Payload....................................................8
3.1 ESP Environmental Considerations............................9
3.2 Keying Material.............................................9
4. Security Considerations........................................9
5. References....................................................10
6. Acknowledgments...............................................11
7. Editors' Addresses............................................12
8. Full Copyright Statement......................................14
1. Introduction
The Encapsulating Security Payload (ESP) [Kent98] provides
confidentiality for IP datagrams by encrypting the payload data to be
protected. This specification describes the ESP use of CBC-mode
cipher algorithms.
While this document does not describe the use of the default cipher
algorithm DES, the reader should be familiar with that document.
[Madson98]
It is assumed that the reader is familiar with the terms and concepts
described in the "Security Architecture for the Internet Protocol"
[Atkinson95], "IP Security Document Roadmap" [Thayer97], and "IP
Encapsulating Security Payload (ESP)" [Kent98] documents.
Furthermore, this document is a companion to [Kent98] and MUST be
read in its context.
1.1 Specification of Requirements
The keywords "MUST", "MUST NOT", "REQUIRED", "SHOULD", "SHOULD NOT",
and "MAY" that appear in this document are to be interpreted as
described in [Bradner97].
1.2 Intellectual Property Rights Statement
The IETF takes no position regarding the validity or scope of any
intellectual property or other rights that might be claimed to
pertain to the implementation or use of the technology described in
this document or the extent to which any license under such rights
might or might not be available; neither does it represent that it
has made any effort to identify any such rights. Information on the
IETF's procedures with respect to rights in standards-track and
standards-related documentation can be found in BCP-11. Copies of
claims of rights made available for publication and any assurances of
licenses to be made available, or the result of an attempt made to
obtain a general license or permission for the use of such
proprietary rights by implementers or users of this specification can
be obtained from the IETF Secretariat.
2. Cipher Algorithms
All symmetric block cipher algorithms share common characteristics
and variables. These include mode, key size, weak keys, block size,
and rounds. All of which will be explained below.
While this document illustrates certain cipher algorithms such as
Blowfish [Schneier93], CAST-128 [Adams97], 3DES, IDEA [Lai] [MOV],
and RC5 [Baldwin96], any other block cipher algorithm may be used
with ESP if all of the variables described within this document are
clearly defined.
2.1 Mode
All symmetric block cipher algorithms described or insinuated within
this document use Cipher Block Chaining (CBC) mode. This mode
requires an Initialization Vector (IV) that is the same size as the
block size. Use of a randomly generated IV prevents generation of
identical ciphertext from packets which have identical data that
spans the first block of the cipher algorithm's blocksize.
The IV is XOR'd with the first plaintext block, before it is
encrypted. Then for successive blocks, the previous ciphertext block
is XOR'd with the current plaintext, before it is encrypted.
More information on CBC mode can be obtained in [Schneier95].
2.2 Key Size
Some cipher algorithms allow for variable sized keys, while others
only allow a specific key size. The length of the key correlates
with the strength of that algorithm, thus larger keys are always
harder to break than shorter ones.
This document stipulates that all key sizes MUST be a multiple of 8
bits.
This document does specify the default key size for each cipher
algorithm. This size was chosen by consulting experts on the
algorithm and by balancing strength of the algorithm with
performance.
+==============+==================+=================+==========+
| Algorithm | Key Sizes (bits) | Popular Sizes | Default |
+==============+==================+=================+==========+
| CAST-128 [1] | 40 to 128 | 40, 64, 80, 128 | 128 |
+--------------+------------------+-----------------+----------+
| RC5 | 40 to 2040 | 40, 128, 160 | 128 |
+--------------+------------------+-----------------+----------+
| IDEA | 128 | 128 | 128 |
+--------------+------------------+-----------------+----------+
| Blowfish | 40 to 448 | 128 | 128 |
+--------------+------------------+-----------------+----------+
| 3DES [2] | 192 | 192 | 192 |
+--------------+------------------+-----------------+----------+
Notes:
[1] With CAST-128, keys less than 128 bits MUST be padded with zeros
in the rightmost, or least significant, positions out to 128 bits
since the CAST-128 key schedule assumes an input key of 128 bits.
Thus if you had a key with a size of 80 bits '3B5D831CFE', it would
be padded to produce a key with a size of 128 bits
'3B5D831CFE000000'.
[2] The first 3DES key is taken from the first 64 bits, the second
from the next 64 bits, and the third from the last 64 bits.
Implementations MUST take into consideration the parity bits when
initially accepting a new set of keys. Each of the three keys is
really 56 bits in length with the extra 8 bits used for parity.
The reader should note that the minimum key size for all of the above
cipher algorithms is 40 bits, and that the authors strongly advise
that implementations do NOT use key sizes smaller than 40 bits.
2.3 Weak Keys
Weak key checks SHOULD be performed. If such a key is found, the key
SHOULD be rejected and a new SA requested. Some cipher algorithms
have weak keys or keys that MUST not be used due to their weak
nature.
New weak keys might be discovered, so this document does not in any
way contain all possible weak keys for these ciphers. Please check
with other sources of cryptography such as [MOV] and [Schneier] for
further weak keys.
CAST-128:
No known weak keys.
RC5:
No known weak keys when used with 16 rounds.
IDEA:
IDEA has been found to have weak keys. Please check with [MOV] and
[Schneier] for more information.
Blowfish:
Weak keys for Blowfish have been discovered. Weak keys are keys that
produce the identical entries in a given S-box. Unfortunately, there
is no way to test for weak keys before the S- box values are
generated. However, the chances of randomly generating such a key
are small.
3DES:
DES has 64 known weak keys, including so-called semi-weak keys and
possibly-weak keys [Schneier95, pp 280-282]. The likelihood of
picking one at random is negligible.
For DES-EDE3, there is no known need to reject weak or
complementation keys. Any weakness is obviated by the use of
multiple keys.
However, if the first two or last two independent 64-bit keys are
equal (k1 == k2 or k2 == k3), then the 3DES operation is simply the
same as DES. Implementers MUST reject keys that exhibit this
property.
2.4 Block Size and Padding
All of the algorithms described in this document use a block size of
eight octets (64 bits).
Padding is used to align the payload type and pad length octets as
specified in [Kent98]. Padding must be sufficient to align the data
to be encrypted to an eight octet (64 bit) boundary.
2.5 Rounds
This variable determines how many times a block is encrypted. While
this variable MAY be negotiated, a default value MUST always exist
when it is not negotiated.
+====================+============+======================+
| Algorithm | Negotiable | Default Rounds |
+====================+============+======================+
| CAST-128 | No | key<=80 bits, 12 |
| | | key>80 bits, 16 |
+--------------------+------------+----------------------+
| RC5 | No | 16 |
+--------------------+------------+----------------------+
| IDEA | No | 8 |
+--------------------+------------+----------------------+
| Blowfish | No | 16 |
+--------------------+------------+----------------------+
| 3DES | No | 48 (16x3) |
+--------------------+------------+----------------------+
2.6 Backgrounds
CAST-128:
The CAST design procedure was originally developed by Carlisle Adams
and Stafford Tavares at Queen's University, Kingston, Ontario,
Canada. Subsequent enhancements have been made over the years by
Carlisle Adams and Michael Wiener of Entrust Technologies. CAST-128
is the result of applying the CAST Design Procedure as outlined in
[Adams97].
RC5:
The RC5 encryption algorithm was developed by Ron Rivest for RSA Data
Security Inc. in order to address the need for a high- performance
software and hardware ciphering alternative to DES. It is patented
(pat.no. 5,724,428). A description of RC5 may be found in [MOV] and
[Schneier].
IDEA:
Xuejia Lai and James Massey developed the IDEA (International Data
Encryption Algorithm) algorithm. The algorithm is described in
detail in [Lai], [Schneier] and [MOV].
The IDEA algorithm is patented in Europe and in the United States
with patent application pending in Japan. Licenses are required for
commercial uses of IDEA.
For patent and licensing information, contact:
Ascom Systec AG, Dept. CMVV
Gewerbepark, CH-5506
Magenwil, Switzerland
Phone: +41 64 56 59 83
Fax: +41 64 56 59 90
idea@ascom.ch
http://www.ascom.ch/Web/systec/policy/normal/exhibit1.html
Blowfish:
Bruce Schneier of Counterpane Systems developed the Blowfish block
cipher algorithm. The algorithm is described in detail in
[Schneier93], [Schneier95] and [Schneier].
3DES:
This DES variant, colloquially known as "Triple DES" or as DES-EDE3,
processes each block three times, each time with a different key.
This technique of using more than one DES operation was proposed in
[Tuchman79].
P1 P2 Pi
| | |
IV->->(X) +>->->->(X) +>->->->(X)
v ^ v ^ v
+-----+ ^ +-----+ ^ +-----+
k1->| E | ^ k1->| E | ^ k1->| E |
+-----+ ^ +-----+ ^ +-----+
| ^ | ^ |
v ^ v ^ v
+-----+ ^ +-----+ ^ +-----+
k2->| D | ^ k2->| D | ^ k2->| D |
+-----+ ^ +-----+ ^ +-----+
| ^ | ^ |
v ^ v ^ v
+-----+ ^ +-----+ ^ +-----+
k3->| E | ^ k3->| E | ^ k3->| E |
+-----+ ^ +-----+ ^ +-----+
| ^ | ^ |
+>->->+ +>->->+ +>->->
| | |
C1 C2 Ci
The DES-EDE3-CBC algorithm is a simple variant of the DES-CBC
algorithm [FIPS-46]. The "outer" chaining technique is used.
In DES-EDE3-CBC, an Initialization Vector (IV) is XOR'd with the
first 64-bit (8 byte) plaintext block (P1). The keyed DES function
is iterated three times, an encryption (Ek1) followed by a decryption
(Dk2) followed by an encryption (Ek3), and generates the ciphertext
(C1) for the block. Each iteration uses an independent key: k1, k2
and k3.
For successive blocks, the previous ciphertext block is XOR'd with
the current plaintext (Pi). The keyed DES-EDE3 encryption function
generates the ciphertext (Ci) for that block.
To decrypt, the order of the functions is reversed: decrypt with k3,
encrypt with k2, decrypt with k1, and XOR the previous ciphertext
block.
Note that when all three keys (k1, k2 and k3) are the same, DES-
EDE3-CBC is equivalent to DES-CBC. This property allows the DES-EDE3
hardware implementations to operate in DES mode without modification.
For more explanation and implementation information for Triple DES,
see [Schneier95].
2.7 Performance
For a comparison table of the estimated speed of any of these and
other cipher algorithms, please see [Schneier97] or for an up-to-date
performance comparison, please see [Bosseleaers].
3. ESP Payload
The ESP payload is made up of the IV followed by raw cipher-text.
Thus the payload field, as defined in [Kent98], is broken down
according to the following diagram:
+---------------+---------------+---------------+---------------+
| |
+ Initialization Vector (8 octets) +
| |
+---------------+---------------+---------------+---------------+
| |
~ Encrypted Payload (variable length) ~
| |
+---------------------------------------------------------------+
1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
The IV field MUST be same size as the block size of the cipher
algorithm being used. The IV MUST be chosen at random. Common
practice is to use random data for the first IV and the last block of
encrypted data from an encryption process as the IV for the next
encryption process.
Including the IV in each datagram ensures that decryption of each
received datagram can be performed, even when some datagrams are
dropped, or datagrams are re-ordered in transit.
To avoid ECB encryption of very similar plaintext blocks in different
packets, implementations MUST NOT use a counter or other low-Hamming
distance source for IVs.
3.1 ESP Environmental Considerations
Currently, there are no known issues regarding interactions between
these algorithms and other aspects of ESP, such as use of certain
authentication schemes.
3.2 Keying Material
The minimum number of bits sent from the key exchange protocol to
this ESP algorithm must be greater or equal to the key size.
The cipher's encryption and decryption key is taken from the first
<x> bits of the keying material, where <x> represents the required
key size.
4. Security Considerations
Implementations are encouraged to use the largest key sizes they can
when taking into account performance considerations for their
particular hardware and software configuration. Note that encryption
necessarily impacts both sides of a secure channel, so such
consideration must take into account not only the client side, but
the server as well.
For information on the case for using random values please see
[Bell97].
For further security considerations, the reader is encouraged to read
the documents that describe the actual cipher algorithms.
5. References
[Adams97] Adams, C, "The CAST-128 Encryption Algorithm",
RFC2144, 1997.
[Atkinson98]Kent, S. and R. Atkinson, "Security Architecture for the
Internet Protocol", RFC 2401, November 1998.
[Baldwin96] Baldwin, R. and R. Rivest, "The RC5, RC5-CBC, RC5-CBC-
Pad, and RC5-CTS Algorithms", RFC 2040, October 1996.
[Bell97] S. Bellovin, "Probable Plaintext Cryptanalysis of the IP
Security Protocols", Proceedings of the Symposium on
Network and Distributed System Security, San Diego, CA,
pp. 155-160, February 1997 (also
http://www.research.att.com/~smb/probtxt.{ps, pdf}).
[Bosselaers]A. Bosselaers, "Performance of Pentium implementations",
http://www.esat.kuleuven.ac.be/~bosselae/
[Bradner97] Bradner, S., "Key words for use in RFCs to indicate
Requirement Levels", BCP 14, RFC 2119, March 1997.
[Crypto93] J. Daemen, R. Govaerts, J. Vandewalle, "Weak Keys for
IDEA", Advances in Cryptology, CRYPTO 93 Proceedings,
Springer-Verlag, pp. 224-230.
[FIPS-46] US National Bureau of Standards, "Data Encryption
Standard", Federal Information Processing Standard (FIPS)
Publication 46, January 1977.
[Kent98] Kent, S. and R. Atkinson, "IP Encapsulating Security
Payload (ESP)", RFC 2406, November 1998.
[Lai] X. Lai, "On the Design and Security of Block Ciphers",
ETH Series in Information Processing, v. 1, Konstanz:
Hartung-Gorre Verlag, 1992.
[Madson98] Madson, C. and N. Dorswamy, "The ESP DES-CBC Cipher
Algorithm With Explicit IV", RFC 2405, November 1998.
[MOV] A. Menezes, P. Van Oorschot, S. Vanstone, "Handbook of
Applied Cryptography", CRC Press, 1997. ISBN 0-8493-
8523-7
[Schneier] B. Schneier, "Applied Cryptography Second Edition", John
Wiley & Sons, New York, NY, 1995. ISBN 0-471-12845-7
[Schneier93]B. Schneier, "Description of a New Variable-Length Key,
64-Bit Block Cipher", from "Fast Software Encryption,
Cambridge Security Workshop Proceedings", Springer-
Verlag, 1994, pp. 191-204.
http://www.counterpane.com/bfsverlag.html
[Schneier95]B. Schneier, "The Blowfish Encryption Algorithm - One
Year Later", Dr. Dobb's Journal, September 1995,
http://www.counterpane.com/bfdobsoyl.html
[Schneier97]B. Scheier, "Speed Comparisons of Block Ciphers on a
Pentium." February 1997,
http://www.counterpane.com/speed.html
[Thayer97] Thayer, R., Doraswamy, N. and R. Glenn, "IP Security
Document Roadmap", RFC 2411, November 1998.
[Tuchman79] Tuchman, W, "Hellman Presents No Shortcut Solutions to
DES", IEEE Spectrum, v. 16 n. 7, July 1979, pp. 40-41.
6. Acknowledgments
This document is a merger of most of the ESP cipher algorithm
documents. This merger was done to facilitate greater understanding
of the commonality of all of the ESP algorithms and to further the
development of these algorithm within ESP.
The content of this document is based on suggestions originally from
Stephen Kent and subsequent discussions from the IPSec mailing list
as well as other IPSec documents.
Special thanks to Carlisle Adams and Paul Van Oorschot both of
Entrust Technologies who provided input and review of CAST.
Thanks to all of the editors of the previous ESP 3DES documents; W.
Simpson, N. Doraswamy, P. Metzger, and P. Karn.
Thanks to Brett Howard from TimeStep for his original work of ESP-
RC5.
Thanks to Markku-Juhani Saarinen, Helger Lipmaa and Bart Preneel for
their input on IDEA and other ciphers.
7. Editors' Addresses
Roy Pereira
TimeStep Corporation
Phone: +1 (613) 599-3610 x 4808
EMail: rpereira@timestep.com
Rob Adams
Cisco Systems Inc.
Phone: +1 (408) 457-5397
EMail: adams@cisco.com
Contributors:
Robert W. Baldwin
RSA Data Security, Inc.
Phone: +1 (415) 595-8782
EMail: baldwin@rsa.com or baldwin@lcs.mit.edu
Greg Carter
Entrust Technologies
Phone: +1 (613) 763-1358
EMail: carterg@entrust.com
Rodney Thayer
Sable Technology Corporation
Phone: +1 (617) 332-7292
EMail: rodney@sabletech.com
The IPSec working group can be contacted via the IPSec working
group's mailing list (ipsec@tis.com) or through its chairs:
Robert Moskowitz
International Computer Security Association
EMail: rgm@icsa.net
Theodore Y. Ts'o
Massachusetts Institute of Technology
EMail: tytso@MIT.EDU
8. Full Copyright Statement
Copyright (C) The Internet Society (1998). All Rights Reserved.
This document and translations of it may be copied and furnished to
others, and derivative works that comment on or otherwise explain it
or assist in its implementation may be prepared, copied, published
and distributed, in whole or in part, without restriction of any
kind, provided that the above copyright notice and this paragraph are
included on all such copies and derivative works. However, this
document itself may not be modified in any way, such as by removing
the copyright notice or references to the Internet Society or other
Internet organizations, except as needed for the purpose of
developing Internet standards in which case the procedures for
copyrights defined in the Internet Standards process must be
followed, or as required to translate it into languages other than
English.
The limited permissions granted above are perpetual and will not be
revoked by the Internet Society or its successors or assigns.
This document and the information contained herein is provided on an
"AS IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING
TASK FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING
BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION
HEREIN WILL NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF
MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.
AI ELI5
On this page
§1Introduction
The Encapsulating Security Payload (ESP) provides confidentiality for IP datagrams by encrypting the payload. RFC 2451 describes how ESP uses CBC-mode cipher algorithms.
It does not cover the default cipher DES, which has its own document (RFC 2405). It is a companion to the ESP specification (RFC 2406) and MUST be read in its context. It assumes familiarity with the IP security architecture (RFC 2401) and the IP security roadmap (RFC 2411).
ESP is a locked envelope for internet packets. This RFC explains how to use one kind of lock, a chained scrambler, inside that envelope.
§2Common Cipher Variables and CBC Mode
All the symmetric block ciphers share the same variables: mode, key size, weak keys, block size and rounds. The RFC illustrates Blowfish, CAST-128, 3DES, IDEA and RC5, but states that any other block cipher may be used with ESP if all those variables are clearly defined.
Every cipher here uses Cipher Block Chaining (CBC). CBC needs an Initialization Vector (IV) the same size as the block size. The IV is XOR'd with the first plaintext block before encryption. For later blocks, the previous ciphertext block is XOR'd with the current plaintext before encryption. A randomly generated IV prevents identical ciphertext from packets with identical data spanning the first block.
Each scrambled piece is mixed with the one before it, so the same words never scramble the same way twice. The first piece is mixed with a random starter.
§3Key Sizes, Rounds and Padding
Key sizes MUST be a multiple of 8 bits. The RFC gives a default for each cipher, chosen by consulting experts and balancing strength with performance. The minimum key size for all of the listed ciphers is 40 bits, and the authors strongly advise against anything smaller.
Two notes matter in practice. A CAST-128 key shorter than 128 bits MUST be zero-padded on the right out to 128 bits. For 3DES the three keys come from three consecutive 64-bit pieces, and implementations MUST take the parity bits into account, since each key is really 56 bits plus 8 parity bits.
All of the algorithms use a block size of eight octets (64 bits), and padding must align the data to be encrypted to an eight octet boundary. The number of rounds MAY be negotiated, but a default MUST always exist. For these ciphers rounds are not negotiable: CAST-128 uses 12 rounds for keys up to 80 bits and 16 for longer keys, RC5 16, IDEA 8, Blowfish 16 and 3DES 48 (16x3).
Longer secret keys are harder to guess, and every cipher chops data into 8-byte chunks, adding filler if needed. Rounds are how many times the scrambling is repeated.
| Algorithm | Key sizes (bits) | Default key | Default rounds |
|---|---|---|---|
| CAST-128 | 40 to 128 | 128 | 12 (key up to 80 bits), 16 (longer) |
| RC5 | 40 to 2040 | 128 | 16 |
| IDEA | 128 | 128 | 8 |
| Blowfish | 40 to 448 | 128 | 16 |
| 3DES | 192 | 192 | 48 (16x3) |
§4Weak Keys and the Ciphers Themselves
Weak key checks SHOULD be performed. A weak key SHOULD be rejected and a new security association requested. The RFC notes that CAST-128 has no known weak keys, RC5 has none known with 16 rounds, IDEA has weak keys, and Blowfish has weak keys that cannot be tested for before the S-box values are generated, though randomly generating one is unlikely.
DES has 64 known weak keys, but for 3DES (DES-EDE3) there is no known need to reject them. However, if the first two or last two 64-bit keys are equal (k1 == k2 or k2 == k3), 3DES is just DES, and implementers MUST reject such keys.
3DES-CBC runs each block through encryption with k1, decryption with k2 and encryption with k3, using the outer chaining technique. If all three keys are equal it is equivalent to DES-CBC, which lets DES-EDE3 hardware work in DES mode. The RFC also gives short backgrounds on each cipher and notes that RC5 is patented and that IDEA is patented in Europe and the United States, with licenses required for commercial use.
Some secret keys are bad choices, like a lock that opens with almost any key. If you get one, throw it away and pick another.
§5The ESP Payload
The ESP payload is the IV followed by the raw ciphertext. The IV field MUST be the same size as the cipher block, which is 8 octets here, and it MUST be chosen at random. Common practice is random data for the first IV and the last block of encrypted data from the previous encryption as the next IV.
Carrying the IV in every datagram means each datagram can be decrypted even if others are dropped or reordered. Implementations MUST NOT use a counter or other low-Hamming-distance source for IVs, to avoid ECB encryption of very similar plaintext blocks in different packets.
For keying, the key exchange must supply at least as many bits as the key size, and the cipher key is taken from the first bits of the keying material. The RFC knows of no problems between these algorithms and other aspects of ESP.
Each packet carries its own random starter value, then the scrambled data. So a lost or shuffled packet never stops the others from being read.
| Field | Name | Size | Value | Description |
|---|---|---|---|---|
IV | Initialization Vector | 8 bytes | 8 octets, chosen at random, same size as the cipher block. |
§6Security Considerations
Implementations are encouraged to use the largest key sizes they can, weighing performance on their hardware and software. Encryption affects both ends of a secure channel, so the cost on the server side counts too. The RFC points to Bellovin's work on the case for random values, and tells readers to consult the documents describing each cipher for further considerations.
Longer keys are safer but slower for both sides, so choose the longest key your computers can comfortably handle.