the vrf IP address is drawn from among the IP addresses associated with interfaces configured for that routing instance. so you meet that requirement.
to enable path checks on the outgoing interface for unicast traffic routed on a VRF routing instance, you may need to include the vrf-mtu-check statement under chassis. [edit chassis] vrf-mtu-check; BR// Masood Blog: http://weblogs.com.pk/jahil/ > Thanks Masood, I have an FE associated with the VRF on the CE/PE router > with a /26 assigned and so I assume I meet that requirement? > > Should path MTU check work within the MPLS cloud ie. Say a link from the > P to PE has a reduced MTU? IS there any general info out there that > would help? I had a look at the RFC. > > Cheers, > Tim. > > -----Original Message----- > From: mas...@nexlinx.net.pk [mailto:mas...@nexlinx.net.pk] > Sent: 27 May 2009 16:28 > To: Hunt,TJ,Tim,DMH2 R > Cc: > Subject: Re: [j-nsp] Path MTU Discovery - MPLS VPN > > path mtu check requires at least one IP address must be associated with > each vrf. If an IP address is not associated with the routing instance, > icmp reply messages cannot be sent. > > BR// > Masood > Blog: http://weblogs.com.pk/jahil/ > > >> Hi, >> >> I've been trying to get Path MTU Discovery working for M/J-series, >> appreciate that the M-series requires vrf-mtu-check <id-10380545.html> > >> setting. Having done this I've tried testing it by launching Pings (DF >> set) with a larger MTU than the link, but I'm getting no response at >> all >> - expecting ICMP response. >> Has anyone any experience of the above? What stops it from working? - >> I have vrf-table-label set on some CE/PE devices. Is my test valid? >> Thanks, >> Tim. >> >> >> >> _______________________________________________ >> juniper-nsp mailing list juniper-nsp@puck.nether.net >> https://puck.nether.net/mailman/listinfo/juniper-nsp >> > > > _______________________________________________ juniper-nsp mailing list juniper-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/juniper-nsp