[c-nsp] Ping with vrf intermittent but ping mpls normal

[Tóth András]

Hi Nanda,

Maybe you have multiple paths between the source and destination? Like
etherchannel or Layer-3 ECMP (route with multiple nexthops on different
links)? That would explain this, possibly the MPLS ping and ping to
Loopback is taking one link, while the previous ping vrf takes a different
one and that link might have some errors on it.

I'd suggest to check the following commands along the path on all possible
physical interfaces both inbound and outbound.
sh interface <x/y>
sh int <x/y> counters error
sh counters error
sh interface <x/y> transceiver detail
sh platform port-asic stats drop <int x/y> (this only works on the 3750)

Best regards,
Andras


On Sat, Dec 8, 2012 at 2:42 AM, nanda pradana ramadhan putra <
animusacre at gmail.com> wrote:

> Hi Oliver,
>
>
> The topology like this JKT-PAJAK-C3750-UPE-01 - MPLS CLOUD
> - JATENG-UNGARAN-R7606-NPE-01 - JATENG-GI.RAWALO-R3845-UPE-01
> The ping is just from gateway vrf on pajak to gateway vrf on rawalo the
> result is always have RTO on it. But ping vrf from pajak to ungaran the
> result is OK. The user on pajak said that they always have packet loss when
> trying to reach user on rawalo. I've investigated it and created new vrf on
> rawalo but the result still same RTO always occur. the other issue when i
> ping using all vrf on rawalo to the other remote site is always RTO. Can be
> it caused by transmission problem? ungaran is upstream path for rawalo.
> Ping p2p interface from ungaran to rawalo always good. The connection is
> using L2DWDM broadcast.
>
> if it caused by ICMP limiter what kind of CoPP that jusl limit icmp ping
> just on vrf? but user also have rto i think it is not caused by CoPP
>
> i try to without mpls just ping using its loopback the result is smooth
> rather ping vrf
>
> JKT-PAJAK-C3750-UPE-01#ping 202.162.222.241 repeat 1000
>
> Type escape sequence to abort.
> Sending 1000, 100-byte ICMP Echos to 202.162.222.241, timeout is 2 seconds:
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
> !!!!!!!!!!!!!!!!!!!!
> Success rate is 100 percent (1000/1000), round-trip min/avg/max = 8/15/118
> ms
>
> Thanks,
> Nanda
>
> On Sat, Dec 8, 2012 at 4:08 AM, Oliver Boehmer (oboehmer) <
> oboehmer at cisco.com> wrote:
>
> >
> >
> > >
> > >i have problem here with ping vrf, when i do ping vrf there is always
> > >packet drop but when i ping using mpls it seems normal. is there some
> > >strange here? i do tracing ping p2p and the result is ok too, but why if
> > >do
> > >ping vrf always have a packet drop
> >
> > you might run into an ICMP rate limiter at the destination (like CoPP or
> > something), which is not effective for MPLS ping (as it isn't based on
> > ICMP)?
> >
> >         oli
> >
> >
> >
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
>