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

[Mahesh Akula <[e-mail]>]

Hi Nitin,

On Tue, Aug 17, 2010 at 12:13 PM, Nitin Bahadur <[e-mail]> wrote:

> Hi Mahesh,
>
> > From this mail thread, can it be assumed that LSP Tracing without IP
> Encapsulation does not work for Associated bidirectional LSP's since there
> may not be a
> > return LSP path from the responding transit node?
>
> For co-routed LSPs, there will always be a return path. For non co-routed
> LSPs, there may not be a return path and non-IP encap might not work.
>
[MA] Okay.

>
> > Also when a node receives LSP Echo request w/o IP encapsulation and
> replymode "not" set to 4, should it drop the packet or should it try to
> respond back with
> >  Malformed Echo request error?
>
>    It should either drop the packet or respond back with the requested
> reply mode.
>

[MA] Responding back with the requested reply mode may not be correct
because while sending the packet, draft explicitly specifies that reply mode
in the lsp ping request MUST be set to 4. If it is not, it means the request
packet is malformed right? If it is suggested to drop the packet in such
case,  I think text may be added in the draft stating this.

>
> Thanks
> Nitin
>

Thanks,
Mahesh