Hi,



From: Filippo Cugini <[EMAIL PROTECTED]>
Date: Wed, 14 Nov 2007 08:02:57 +0100
To: <[EMAIL PROTECTED]>
Subject: [Pce] Liveness in draft-ietf-pce-monitoring-00.txt

Hi,
 
just a consideration on the Liveness information in the pce-monitoring
draft.
 
In case of specific monitoring request:
if one PCE involved in the path computation chain is not alive the
computation should fail and, rather than a monitoring object, maybe a
specific error could be more appropriate.
 
In case of general monitoring requests, maybe the discovery mechanism could
be more appropriate for such information.
 
but maybe I missed something. Could you please comment on that?

JP> Let¹s consider the case of PCC1 ---- PCE1 ---- PCE2

Few comments:

1. PCC1 may of course use a timer upon sending a request after the
expiration of which it will resend its request if no reply has been received
(which covers the case where PCE1 or PCE2 fail). This ID is not aimed at
being used for that purpose.
2. If PCE1 dies, indeed PCC1 may discover PCE¹s failure thanks to the IGP if
dynamic PCE discovery is in use (which may not be always the case).
Furthermore PCEP could be used to detect a PCE failure. But again the aim of
this ID is not to solve that specific issue.

The aim of this ID is to be used by the network operator or an NMS to
³monitor² a PCE chain. For example, the user may want to check the liveness
of the PCE chain at any point in time, which can be used with PCMonReq
message, without having to send an actual path computation request.

Thanks.

JP.

 
Thank you
   Filippo


_______________________________________________
Pce mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/pce

_______________________________________________
Pce mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/pce

Reply via email to