RE: Question on Restart Procedure - Restart Time

["Ong, Lyndon" <LyOng@ciena.com>]

Hi Suresh,

This was a source of confusion back in December, at one
point it looked like an "infinite" value was going to be
defined for Restart Time rather than Recovery Time.

Apparently this was not done, can someone recall what the
resolution was?  I'd check the archive but this seems to
be down at the moment...

Cheers,

Lyndon 

-----Original Message-----
From: Suresh Katukam [mailto:skatukam@cisco.com]
Sent: Wednesday, April 10, 2002 5:25 PM
To: Zhi-Wei Lin
Cc: ccamp@ops.ietf.org
Subject: Re: Question on Restart Procedure - Restart Time



Zhi,

This only allows the local node to wait for longer. If I want the other 
node to wait
forever, ( I do not want to depend on the other node's local policy ), I 
should be
able to indicate to the other node that - Please wait forever or Until Hello
established and Recovery timer expires.

Thanks,
Suresh

At 07:20 PM 4/10/2002 -0400, Zhi-Wei Lin wrote:
>Hi Suresh,
>
>In the GMPLS RSVP-TE document, Section 9.3, first paragraph, last sentence:
>
>A node MAY wait longer based on local policy or configuration information.
>
>Is this what you are looking for?
>
>Zhi
>
>
>Suresh Katukam wrote:
>
>>
>>>Ping and others,
>>>
>>>Is it possible to define an infinite value for Restart time so that one 
>>>may not
>>>have to remove circuits as soon as Restart timer expires? This is more
valid
>>>in SONET/SDH cases where a control channel may be down for a long enough
>>>time that is difficult to guess it up front..
>>>
>>>In SONET/SDH cases, nodes should delete circuits in these conditions
>>>
>>>- explicitly removed by customer
>>>- Recovery Timer expires
>>>
>>>other conditions ( restart timer expires, soft-state time out ) should 
>>>be given
>>>optional. there should be values or procedures defined in these other
cases
>>>such that the node does not remove LSPs.
>>>
>>>Thanks,
>>>Suresh
>>
>>
>