The MPLS-OPS Archive

[Robert Raszuk <raszuk@cisco.com>]

> I came across literature which mentioned that, e.g., when TTL reaches 0 on
> P2 (in the network shown above) for a traceroute probe initiated from PE1 to
> CE2, the 'ICMP TTL Expired' travel to PE2 first and then back to PE1 as P2
> is in the middle of LSP tunnel and does not know how to send the error back
> to PE1. Is that a valid scenario? if yes, which standard lists this
> scenario?

Yes this is valid.

See: http://www.ietf.org/proceedings/01mar/I-D/mpls-icmp-02.txt

   The current memo proposes extensions to ICMP that permit an LSR to 
   append MPLS label stack information to ICMP messages. ICMP messages 
   regarding MPLS encapsulated datagrams SHOULD include the MPLS label 
   stack, as it arrived at the router that is sending the ICMP message. 
   The ICMP message MUST also include the IP header and leading payload 
   octets of the original datagram.  

R.

> "Dhananjay S. Patki" wrote:
> 
> Hi,
> 
> I am looking for how ping & traceroute works in MPLS VPN scenarios. Any
> references to concerned standards/drafts will be appreciated.
> 
> (VPNa) CE1 ---- PE1 ---- P1 ---- P2 ---- PE2 ---- CE2 (VPNa)
> 
> I came across literature which mentioned that, e.g., when TTL reaches 0 on
> P2 (in the network shown above) for a traceroute probe initiated from PE1 to
> CE2, the 'ICMP TTL Expired' travel to PE2 first and then back to PE1 as P2
> is in the middle of LSP tunnel and does not know how to send the error back
> to PE1. Is that a valid scenario? if yes, which standard lists this
> scenario?
> 
> Thanks,
> Dhananjay
> 
> -------
> The MPLS-OPS Mailing List
> Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
> Archive: http://www.mplsrc.com/mpls-ops_archive.shtml

-------
The MPLS-OPS Mailing List
Subscribe/Unsubscribe:  http://www.mplsrc.com/mplsops.shtml
Archive: http://www.mplsrc.com/mpls-ops_archive.shtml