Re: [mpls] Question on scope of draft-frost-mpls-loss-delay-00

[Stewart Bryant <[e-mail]>]

On 20/12/2010 03:34, Greg Mirsky wrote:
> Dear Dan and Stewart,
> bullet two in Introduction and Section 2.7.1 Type of Channels list 
> only link (section, LSP or a PW as channels over which LM and DM could 
> be conducted. But the Section 2.7.4 Intermediate Nodes describes 
> operation of LM and DM by using TTL exception method. Though LM and DM 
> operation does not differentiate whether querier and addressee are MEP 
> or MIP reference to use of TTL implies to LM/DM directed at MIP. I 
> view it as case of Segment. Do you see that querier might be MIP as 
> well or LM/DM on generic Segment can be performed only with SPME?
>
> Regards,
> Greg
In MPLS  I can see no reason why one could not make the measurement 
between any two nodes on the LSP using the TTL method.

Within the MPLS-TP of this, the convention is to only initiate the OAM 
exchange at an LSP endpoint, thus requiring an SPME to monitor the 
traffic between a pair of midpoint LSRs. We may need to add some text on 
this in the MPLS-TP for loss-delay.

- Stewart