Re: [PWE3] Comments on draft-jc-pwe3-mpls-tp-static-checking-00.txt

[Lizhong Jin<[e-mail]>]

Hi Sri,
See inline please. Thank you for the comments.

Lizhong


> ------------------------------
> 
> Message: 2
> Date: Wed, 28 Mar 2012 07:30:19 -0700
> From: Sriganesh Kini <[e-mail]>
> To: "[e-mail]" <[e-mail]>
> Subject: [PWE3] Comments on
>    draft-jc-pwe3-mpls-tp-static-checking-00.txt
> Message-ID:
>    <[e-mail]>
> Content-Type: text/plain; charset="utf-8"
> 
> Comments
> 
> 1. My first thought on looking at this - what is the relationship of 
this
> to a signaling protocol (LDP) ? Is this trying to replicate that
> functionality or provide som new functionality ?
[Lizhong] this mechanism is to only check the configuration, after 
finished, the mechanism will be stopped. So it is a checking tool, not a 
signaling protocol. The checking procedure is part of same as that does in 
signaling protocol (LDP), but there is also exception, e.g, control word 
checking.

> 
> 2. Is this checking intended to be used as operator initiated or 
proactive
> checking ?
[Lizhong] depends on the implementation. In the draft, it is suggested to 
be trigged basically by configuration changing. It could be also triggered 
by a CLI command.
> 
> 3. The term "corresponding PSN tunnel" has been used. I guess this means
> the PSN tunnel to which the PW is nailed to (?) 
[Lizhong] yes

> However there is no receive
> check to ensure that the received msg is indeed on a tunnel that is
> carrying the PW. Even if there was, what is the need to map this to the
> 'corresponding' PSN tunnel ?
[Lizhong] GAP should carry LSP Identifier in the message, and then could 
check if the LSP Identifier is the corresponding PSN tunnel. The draft is 
lack of this content, and would be added in new version. Only when its 
corresponding PSN Tunnel is UP, the PW could be UP, so we use PW's 
underlying PSN tunnel to transmit parameters. 

> 
> 
> 4. If just the in-label is sent can this be used as a label 
advertisement
> mechanism ?
[Lizhong] yes, could do that, and is only used to advertise label, nothing 
else (e.g, status). In that case, the static PW could be configured only 
with in-label, to reduce configuration burden.

> 
> 
> 
> -- 
> - Sri
> ------------------------------
>