Re: [mpls-tp] Comments on draft-nitinb-mpls-tp-on-demand-cv-00

[Mach Chen <[e-mail]>]

Hi Nitin,

Thanks for your reply!

See my reply inline...
--------------------------------------------------
From: "Nitin Bahadur" <[e-mail]>
Sent: Friday, June 25, 2010 1:56 AM
To: "'Mach Chen'" <[e-mail]>om>; <[e-mail]>
Subject: RE: [mpls-tp] Comments on draft-nitinb-mpls-tp-on-demand-cv-00

> Hi Mach,
>
>   Very good example. See more below.
>
>> IMHO, even if in non-error cases(for associated bidirectional
>> LSP), the MIPs response may not be able to send the response,
>
>> see the figure below:
>> A---B---C---D---E
>>  \             /
>>   F----G------H
>> A associated bidirectional LSP combined with two
>> unidirectional LSPs( LSP1:
>> A->B->C->D->E, and LSP2: E->H->G->F->A), when the trace
>> request reach at
>> A->B->C->D->B,
>> how does node B send the response?
>
> The above case is probably not how things get deployed in transport
> networks (transport folks can correct me if I am wrong). Transport 
> networks
> typically have bi-directional physical paths.

IMHO, it's better for transport network to have bi-directioanl phyical 
paths, but it could not exclude the situation where the two directions are 
diverse(at least partially diverse), and this is one of the reasons that 
associated bidrectional LSP required in MPLS-TP.

>
> In any case, there are 2 ways for B to send a response back,
> 1) B sends the response back directly to A via IP or some other 
> encapsulation.
> 2) B tunnels the packet to E and sends the response back.
>
> Option 2 is not practical & insufficient, because if there is a fault at 
> C, then the echo
> response will never make it back...giving the impression that the fault is 
> at B.
> Adding the complexity to tunnel the packet to E and then back seems too 
> cumbersome
> and I don't see a real business case/requirement for the same.
>
> Option 1 will not work if there is no IP/MPLS path back to A from B....but 
> I consider that
> more of a network design issue....

I basically agree with you analyse, so for associated bidirectional LSP, the 
operation is almost the same unidirectional LSP.

One more question: does the tracing result of an associated bidirectional 
LSP inlcude the reverse path, or just one direction? if incluing reverse 
direction, how to process and present the result?

Best regards,
Mach