The MPLS WG Archive

[George Swallow <swallow@cisco.com>]

Dave -

> ......So I'll ask a simple technical question to try and spark some
> discussion after the fact. If I have an LSR popping the stack in error, I
> suspect this will not be detected by an LSR testing itself. As PHP is
> considered acceptable behavior, the downstream LSR would NOT observe that
> the PING packet on which TTL expired did not arrive as a labelled packet, it
> simply looks at the FEC TLV and goes ***OK, it looks reasonable***. There is
> no guarantee that the LSR downstream of the defective pop would catch this
> either via its use of LSR self test mechanisms as the path into the
> defective LSR is assuredly different than that of real traffic.
> 
> IMHO if my LSRs are generating a ping transaction for every ILM entry in my
> network (which is a s***load of a lot of pings no matter how slowly I run
> this stuff), am I really accomplishing anything if discontinuties in the
> label switched path cannot be detected by this mechanism? 


Excellent point you raise.  One that I've been worried about.  I've
sent some new procedures and error conditions to my co-authors on
lsp-ping.  One change that I have requested is is that you must check
whether the protocol that would have distributed the FEC you have is
actually associated with the interface that you received the packet
on.  And that that protocol distributed a label (including NULL) for
that FEC.  Note, I say associated since and LDP session can pertain to
multiple interfaces and BGP can pertain to all core facing interfaces.

Another change, intended to help diagnose the problem, is that when
you do a trace with a Downstream Mapping TLV, you MUST include null
labels in the stack. That way, if you reach the PHP router and it is
doing a PoP and forward because it is the end of the LSP (got no label
from it's neighbor) it will send back an empty stack, whereas if it
was appending an implicit NULL label (still a PHP) it would send back
a stack with just the null label as the first entry.

...George

========================================================================
George Swallow             Cisco Systems                  (978) 936-1398
                           1414 Massachusetts Avenue
                           Boxborough, MA 01719