Re: [Detnet] Packet Ordering Function (draft-varga-detnet-pof-00.txt)

[Stephan Kehrer <[e-mail]>]

Hi Greg,

Looking at the scenario you describe I think packets arriving in the order 100, 103, 102, 101 cannot happen due to an error on the transmission path if you are using only two paths. They might happen if the sender is failing and sends out the packets in the wrong order to begin with or if you are using more than two paths for redundancy and have failures on multiple transmission paths in a very specific order.
That being said, even then the transmission order 100, 101, 103, 102 should not be something you get from the POF in my opinion. If the packets are arriving and are buffered the way you describe, I think the following is what should happen (changes to what you had in your mail marked in red):

  *   POFLastSent = 100
  *   a packet with seq_num = 103 is received and placed into the buffer for the duration of POFMaxDelay
  *   next received packet has seq_num = 102. It is also moved into the buffer for the duration of POFMaxDelay
  *   finally, before all delayed packets transmitted out of the buffer, a packet with seq_num = 101 is received and is forwarded right away. This sets POFLastSent = 101.
  *   the packet with seq_num=102 is forwarded because it now fulfills the condition ‘OR its seq_num becomes equal to"POFLastSent + 1"’. This sets POFLastSent = 102.
  *   the packet with seq_num=103 is forwarded because it now fulfills the condition ‘OR its seq_num becomes equal to"POFLastSent + 1"’. This set POFLastSent = 103.
  *   Packets arriving after that are forwarded without delay if they are next in sequence, and delayed if they are out of sequence as described in the POF algorithm.
Regards,
Stephan
--
Mit freundlichen Grüßen | With kind regards

Stephan Kehrer | Senior Architect – CTO Office | Hirschmann Automation and Control GmbH
Stuttgarter Straße 45-51 | 72654 Neckartenzlingen | Germany
Phone: [phone] - 1035 | Mobile: [phone] | Fax: [phone] - 1600
[e-mail]<mailto:[e-mail]> | www.beldensolutions.com<http://www.beldensolutions.com/> | www.blog.beldensolutions.com<http://www.blog.beldensolutions.com/>

Hirschmann Automation and Control GmbH, Neckartenzlingen
Register Court: Stuttgart, Trade Register No.: HRB 225927
VAT No.: DE 814 212 604
Managing Director: Brian Anderson, Wolfgang Schenk, Brian Lieser



Von: detnet <[e-mail]> Im Auftrag von Greg Mirsky
Gesendet: Samstag, 24. April 2021 03:04
An: Balázs Varga A <balazs.a.varga=[e-mail]>
Cc: [e-mail]
Betreff: [EXTERNAL] Re: [Detnet] Packet Ordering Function (draft-varga-detnet-pof-00.txt)


External Message:Use caution before opening links or attachments
Hi Balazs,
thank you for the well-designed mechanism of one of the key DetNet functions.
I have a scenario and would appreciate your consideration if you think it is a plausible one:

  *   assume that POFLastSent = 100
  *   a packet with seq_num = 103 is received and placed into the buffer for the duration of POFMaxDelay
  *   next received packet has seq_num = 102. It is also moved into the buffer for the duration of POFMaxDelay
  *   finally, before all delayed packets transmitted out of the buffer, a packet with seq_num = 101 is received and is forwarded right away
  *   the delay timer expires on the packet with seq_num=103. It is forwarded and  POFLastSent = 103
  *   next, the delay timer expires on the packet with seq_num=102. It, in turn, is forwarded and  POFLastSent = 102
  *   from now, based on my understanding of the described POF algorithm, every packet will be delayed even if there are no re-ordered packets. For example, the sequence of arriving packets is as 104, 105, 106, ...
Much appreciate your opinion if the scenario, i.e., a sequence like 100, 103, 102, 101, is a realistic one.

Regards,
Greg

On Mon, Apr 19, 2021 at 10:39 PM Balázs Varga A <balazs.a.varga=[e-mail]<mailto:[e-mail]>> wrote:
Hi,

A draft on Packet Ordering Function (POF) algorithms were uploaded.
Comments and views are highly appreciated.

Thanks
Bala'zs (and the authors)

-----Original Message-----
From: [e-mail]<mailto:[e-mail]> <[e-mail]<mailto:[e-mail]>>
Sent: Tuesday, April 20, 2021 7:14 AM
To: Balázs Varga A <[e-mail]<mailto:[e-mail]>>; Janos Farkas <[e-mail]<mailto:[e-mail]>>; Balázs Varga A <[e-mail]<mailto:[e-mail]>>; Janos Farkas <[e-mail]<mailto:[e-mail]>>; Stephan Kehrer <[e-mail]<mailto:[e-mail]>>; Stephan Kehrer <[e-mail]<mailto:[e-mail]>>; Tobias Heer <[e-mail]<mailto:[e-mail]>>; Tobias Heer <[e-mail]<mailto:[e-mail]>>
Subject: New Version Notification for draft-varga-detnet-pof-00.txt


A new version of I-D, draft-varga-detnet-pof-00.txt has been successfully submitted by =?utf-8?q?Bal=C3=A1zs_Varga?= and posted to the IETF repository.

Name:           draft-varga-detnet-pof
Revision:       00
Title:          Deterministic Networking (DetNet): Packet Ordering Function
Document date:  2021-04-19
Group:          Individual Submission
Pages:          10
URL:            https://www.ietf.org/archive/id/draft-varga-detnet-pof-00.txt<https://urldefense.com/v3/__https:/www.ietf.org/archive/id/draft-varga-detnet-pof-00.txt__;!!B_EDKQmmvsU!7vmDpLfsfmmLuG9tQHwCrsPqlNqL_fA88lGK7awH9s4iu2A-gZ-KVq6uPhKWdzZ8c98$>
Status:         https://datatracker.ietf.org/doc/draft-varga-detnet-pof/<https://urldefense.com/v3/__https:/datatracker.ietf.org/doc/draft-varga-detnet-pof/__;!!B_EDKQmmvsU!7vmDpLfsfmmLuG9tQHwCrsPqlNqL_fA88lGK7awH9s4iu2A-gZ-KVq6uPhKWJABOkBc$>
Htmlized:       https://datatracker.ietf.org/doc/html/draft-varga-detnet-pof<https://urldefense.com/v3/__https:/datatracker.ietf.org/doc/html/draft-varga-detnet-pof__;!!B_EDKQmmvsU!7vmDpLfsfmmLuG9tQHwCrsPqlNqL_fA88lGK7awH9s4iu2A-gZ-KVq6uPhKWE_vUf0k$>
Htmlized:       https://tools.ietf.org/html/draft-varga-detnet-pof-00<https://urldefense.com/v3/__https:/tools.ietf.org/html/draft-varga-detnet-pof-00__;!!B_EDKQmmvsU!7vmDpLfsfmmLuG9tQHwCrsPqlNqL_fA88lGK7awH9s4iu2A-gZ-KVq6uPhKW2P84vhs$>


Abstract:
   Replication and Elimination functions of DetNet [RFC8655] may result
   in out-of-order packets, which may not be acceptable for some time-
   sensitive applications.  The Packet Ordering Function (POF) algorithm
   described herein enables to restore the correct packet order when
   replication and elimination functions are used in DetNet networks.




Please note that it may take a couple of minutes from the time of submission until the htmlized version and diff are available at tools.ietf.org<https://urldefense.com/v3/__http:/tools.ietf.org__;!!B_EDKQmmvsU!7vmDpLfsfmmLuG9tQHwCrsPqlNqL_fA88lGK7awH9s4iu2A-gZ-KVq6uPhKWhEGXwQU$>.

The IETF Secretariat


_______________________________________________
detnet mailing list
[e-mail]<mailto:[e-mail]>
https://www.ietf.org/mailman/listinfo/detnet<https://urldefense.com/v3/__https:/www.ietf.org/mailman/listinfo/detnet__;!!B_EDKQmmvsU!7vmDpLfsfmmLuG9tQHwCrsPqlNqL_fA88lGK7awH9s4iu2A-gZ-KVq6uPhKWTJ53qCw$>

**********************************************************************
DISCLAIMER:
Privileged and/or Confidential information may be contained in this message. If you are not the addressee of this message, you may not copy, use or deliver this message to anyone. In such event, you should destroy the message and kindly notify the sender by reply e-mail. It is understood that opinions or conclusions that do not relate to the official business of the company are neither given nor endorsed by the company. Thank You.