RFC 2406 (IPSec ESP) specifies out of order handling. It describes the use of
a sliding window scheme, with a minimum size of 32 and a recommended default
size of 64. So clearly any conforming implementation will handle a fair amount
of reordering; how much more than the minimum would be up to implementation
choice, or perhaps set by an adjustable parameter. For this reason I would not
describe IPSec as “sensitive to reordering.
paul
Internal Use - Confidential
From: Ingemar Johansson S <[email protected]>
Sent: Tuesday, March 24, 2026 5:15 AM
To: [email protected]
Cc: Greg White <[email protected]>; Chris Box <[email protected]>;
[email protected]; [email protected]; [email protected];
Ingemar Johansson S <[email protected]>
Subject: [IPsec] Re: IPSEC and packet reordering
Hi
I try and post this again. I think that we have a fairly good understanding
about the implications of out of sequence delivery on the transport protocol
level. What is lacking is an understanding on how IP Sec can handle out of
sequence delivery and what the pain limits are. So any kind of input that helps
us advance the https://datatracker.ietf.org/doc/draft-white-intarea-reordering/
draft is welcome.
Regards
Ingemar
From: Ingemar Johansson S
<[email protected]<mailto:[email protected]>>
Sent: Thursday, 5 March 2026 10:40
To: [email protected]<mailto:[email protected]>
Cc: Greg White <[email protected]<mailto:[email protected]>>; Chris Box
<[email protected]<mailto:[email protected]>>;
[email protected]<mailto:[email protected]>;
[email protected]<mailto:[email protected]>;
[email protected]<mailto:[email protected]>; Ingemar Johansson
S <[email protected]<mailto:[email protected]>>
Subject: Re : IPSEC and packet reordering
Hi
I re-listened in on the presentation by Chris Box at the IPSECME session at
IETF-124
https://www.youtube.com/watch?v=ZwfypFEEcYY&t=6866s
The title was “Proposal for updates to Guidance on Packet Reordering”, with the
accompanying IETF draft
https://datatracker.ietf.org/doc/draft-white-intarea-reordering/
The topic of packet reordering has recently been brought up in the 3GPP RAN2
working group where HoL blocking due to link layer retransmission and and
resequencing in 3GPP is seen as an issue.
https://www.3gpp.org/ftp/tsg_ran/WG2_RL2/TSGR2_133/Docs/R2-2600101.zip
Removing or modifying the resequencing in cellular networks would remedy the
issue presented in the document. The question is how IPSEC in particular would
manage this ?.
The discussion at the end of the IPSECME session gave some insight but I am not
sure that it gave any conclusive message. So please comment on what can be
recommended/tolelated in terms of packet reordering from an IP SEC perspective.
Regards
Ingemar + other authors of draft-white-intarea-reordering
=================================
Ingemar Johansson M.Sc.
Master Researcher
Ericsson Research
GFTL ER NAP NCM Netw Proto & E2E Perf
Laboratoriegränd 11
977 53, Luleå, Sweden
+46-73 078 3289
[email protected]<mailto:[email protected]>
www.ericsson.com<http://www.ericsson.com/>
Experience is merely the name
men gave to their mistakes.
Oscar Wilde
=================================
_______________________________________________
IPsec mailing list -- [email protected]
To unsubscribe send an email to [email protected]
