I remember last time I also encounter such issue caused by microburst, we change the input q size to 2000 from default 70. And also I remember we can check the buffer loss is how many packets through show buffer command.
Thanks and regards, Xu Hu On 20 Jun, 2012, at 6:09, Nick Hilliard <n...@foobar.org> wrote: > On 19/06/2012 21:37, sledge...@gmail.com wrote: >> I read a reply from Waris saying that microburst can cause drops when >> traffic arrives from high speed interfaces, he suggested increasing the >> queue-limit to 492 KB, the box only has a total of 44MB for packet >> buffer and I would like to monitor how much of this total is in use at >> any one time, is there an snmp oid that would show me this usage or any >> other way I can monitor this. >> >> Maybe One for Waris if he still monitors the list or anybody else who >> has come across this situation before. > > Eh, difficult to monitor. Remember that this is a dynamic buffer: in the > time that it takes to send and receive an snmp query from a busy host, a > 492K buffer might have been filled and drained several times. > > Nick > _______________________________________________ > cisco-nsp mailing list cisco-nsp@puck.nether.net > https://puck.nether.net/mailman/listinfo/cisco-nsp > archive at http://puck.nether.net/pipermail/cisco-nsp/ _______________________________________________ cisco-nsp mailing list cisco-nsp@puck.nether.net https://puck.nether.net/mailman/listinfo/cisco-nsp archive at http://puck.nether.net/pipermail/cisco-nsp/