Re: [Isis-wg] Question about Purge LSP in RFC 1142

[mike shand <[e-mail]>]

Theodore.Wang wrote:

Hi, all

 

I have a question about Purge LSP in RFC 1142.

In RFC 1142, I find the following lines:

You should not be paying any attention to RFC 1142. It is a copy of DP 10589 which is a version of ISO/IEC 10589 several stages before its final standardisation and differs from ISO/IEC 10589 is several important respects.

If you want to know what the standard says, get a copy of ISO/IEC 10589:2002 from

http://standards.iso.org/ittf/PubliclyAvailableStandards/" rel="nofollow">http://standards.iso.org/ittf/PubliclyAvailableStandards/

 

e)An Intermediate system receiving a Link State PDU with an incorrect LSP Checksum or with an invalid PDU syntax shall
1)log a circuit notification, corruptedLSPReceived,
2)overwrite the Checksum and Remaining Lifetime with 0, and
3)treat the Link State PDU as though its Remaining Lifetime had expired

These lines define the way how to cope with the LSP with incorrect checksum, that is to flood the LSP's header with zero remaining lifetime.

In fact this behaviour has been changed in the second edition and now reads

e) An Intermediate system receiving a Link State PDU with an incorrect LSP Checksum or with an invalid PDU syntax shall
1) generate a corruptedLSPReceived circuit event,
2) discard the PDU.

see also RFC 3719 section 8, which reads

8.  Purging Corrupted PDUs

   While ISO 10589 requires in section 7.3.14.2 e) that any LSP received
   with an invalid PDU checksum should be purged, this has been found to
   be disruptive.  Most implementations today follow the revised
   specification, and simply drop the LSP.

   In ISO 10589:2002 [1], Section 7.3.14.2, it states:

      (e)  An Intermediate system receiving a Link State PDU with an
           incorrect LSP Checksum or with an invalid PDU syntax SHOULD

           1) generate a corruptedLSPReceived circuit event,

           2) discard the PDU.

 

But in fact, in a huge network, it is hard to locate the router that generate the Purge LSP, and which is the situation I'm facing. A Corrupted LSP Storm is happening in the network and it lacks a efficient way to troubleshooting because we can only use "ignore checksum error" on every router to find the one that flooded the Purge LSP.

It is for this reason (the purge storm) that the specification was changed!

 

Therefore, is it possible to attach a new TLV in Purge LSP that defines the originator?

No.

 

Thanks in advance!

_______________________________________________
Isis-wg mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/isis-wg" rel="nofollow">https://www1.ietf.org/mailman/listinfo/isis-wg
  

_______________________________________________
Isis-wg mailing list
[e-mail]
https://www1.ietf.org/mailman/listinfo/isis-wg