[j-nsp] Strange issue

Post by james list
Dear experts
I'm wondering if you can provide any hints/help on this problem.
We experienced a strange issue in reaching the remote devices (servers) and
perforiming bulk snmp walk, instead direct object query was working fine.

Sounds like a temporary MTU problem to me ...

Post by james list
After an entire nigth of issues (22pm to 6am), the problem disappeared
alone...

... on a backup path of that carrier, which was used during this time
and *should* have protected your circuit but did not deliver the full
MTU you require. When the primary path came back up, your service
return to normal.

Have the carrier confirm the reroute and ask to check the MTU on the
backup path.

Post by james list
I think the carrier had some problem but I'm not able to prove it.
Have you never seen this kind of issue ?

Always check the MTU in those cases and consider implementing it as an
additional check in your monitoring system.

lukas
_______________________________________________
juniper-nsp mailing list juniper-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/juniper-nsp

Saku Ytti

2018-09-10 22:16:14 UTC

Post by james list
We experienced a strange issue in reaching the remote devices (servers) and
perforiming bulk snmp walk, instead direct object query was working fine.

Sounds like a temporary MTU problem to me ...

Please elaborate. Bulk walk does bulk get 'OID X and next Y OIDs'.
Specific get is 'OID X'. So what OP is proposing sounds like smaller
UDP datagrams didn't pass, but larger did. My doge coins would be on
edge policer limiting small UDP packets.

--
++ytti
_______________________________________________
juniper-nsp mailing list juniper-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/juniper-nsp

Saku Ytti

2018-09-10 22:19:51 UTC

Oh I think I misunderstood OP. Yes, sounds like larger packets were
impacted smaller were not.

Post by james list
We experienced a strange issue in reaching the remote devices (servers) and
perforiming bulk snmp walk, instead direct object query was working fine.

Sounds like a temporary MTU problem to me ...

--
++ytti
_______________________________________________
juniper-nsp mailing list juniper-***@puck.nether.net
https://puck.nether.net/mailman/listinfo/juniper-nsp

james list

2018-09-12 17:38:34 UTC

Thanks Saku/Lukas
Investigation still on going I will let you know if something is found.

Cheers

Post by Saku Ytti
Oh I think I misunderstood OP. Yes, sounds like larger packets were
impacted smaller were not.

Post by james list
We experienced a strange issue in reaching the remote devices

(servers) and

Post by james list
perforiming bulk snmp walk, instead direct object query was working

fine.