Re: Detection and Recovery Time

[Dimitri.Papadimitriou@alcatel.be]

hi,

gmpls or more generally the control plane is not strictly speaking 
involved in the fault detection this is left to the data plane (thus 
if the detection time takes for instance, 1 ms when gmpls is not 
available it will take 1 ms when it is) the main improvment from 
the control plane viewpoint (and vertical information exchange b/w
the data and the control plane) is to decrease the time needed to 
make this information available at the control plane level (i guess 
this is what you mean by the triggering ?)

in the shared (path) protection case you've mentioned here gmpls 
does not require to replicate the in-band signalling when the 
data plane delivers but is to be used for syncing actions b/w 
edge points for instance, but also control further reversion 
actions (ie normalization), re-provisioning, etc. 

when used the expected time performance are on the order of
100ms since recovery path are fully provisioned (also keep in 
mind the numerous dimensions on which this process relies)

hope this clarifies a bit, additional information can be found
in the following i-d's they were discussed over past meetings
and are to be considered (as usual) as ongoing efforts:

http://www.ietf.org/internet-drafts/draft-ietf-ccamp-gmpls-recovery-terminology-00.txt
http://www.ietf.org/internet-drafts/draft-papadimitriou-ccamp-gmpls-recovery-analysis-02.txt
http://www.ietf.org/internet-drafts/draft-bala-gmpls-recovery-functional-00.txt

feedback is always welcome,
- dimitri.

> Blatt Marcelo wrote:
> 
> Hi,
> 
> I would like to know your opinion regarding GMPLS fault detection and
> recovery time. Ideally, one would like to get SDH times (i.e. 60 ms,
> 10 ms for detection and 50 ms for switchover) but this seems to be
> achievable only in the case of dedicated 1+1 (SNCp) path protection.
> In the Shared Protection case, where signaling is involved, this seems
> to be a very hard to be achieved. What should be a realistic switching
> time? (in what scale 100 ms, 100's ms or seconds).
> 
> I am also wondering regarding the triggering, what should be a
> reasonable integration time?
> 
> Thank you in advance,
> 
> Marcelo Blatt

-- 
Papadimitriou Dimitri 
E-mail : dimitri.papadimitriou@alcatel.be 
Private: http://www.rc.bel.alcatel.be/~papadimd/index.html
E-mail : dpapadimitriou@psg.com
Public : http://psg.com/~dpapadimitriou/
Address: Fr. Wellesplein 1, B-2018 Antwerpen, Belgium
Phone  : Work: +32 3 2408491 - Home: +32 2 3434361