The MPLS WG Archive

["Adrian Farrel" <adrian@olddog.co.uk>]

Hi,

In Paris we asked two questions about P2MP LSP Ping (now
draft-ietf-mpls-p2mp-lsp-ping-00.txt)

1. Is there a requirement to be able to ping a subset of the total set of
leaves?

Currently the draft allows you to ping one or all of the leaves. If you
ping all of the leaves you may specify a jitter to reduce the stress on
the ingress LSR and its surrounding network.

An earlier version of the draft included the ability to specify a subset
of leaves that were being pinged. We removed this feature because we felt
that it placed significant processing overhead leaves during ping
processing and on branches during traceroute processing.

We would like to hear from you if you believe that we should reintroduce
this feature.

2. Should we include support for multicast-LDP in this I-D?

It is clear that many of the features of P2MP LSP Ping will be the same
for P2MP MPLS-TE and mLDP. However, several features will be considerably
different, not least the fact that the ingress of an mLDP LSP is unlikely
to know the identities of the egresses. This might put ping for mLDP LSPs
into the same arena as ping for multicast IP.

Given these considerations and the fact that P2MP MPLS-TE is considerably
more advanced through the specification process than mLDP, should we aim
to include both in this I-D, or should we separate them?

Thanks for any opinions.

Adrian


_______________________________________________
mpls mailing list
mpls@lists.ietf.org
https://www1.ietf.org/mailman/listinfo/mpls