[Linux-cluster] Anyone had any luck changing the NETMTU +1500 in cluster.conf?

Abraham Alawi a.alawi at auckland.ac.nz
Sun Jan 17 22:17:47 UTC 2010

I'm using AoE (CoRAID) as backend storage for GFS2 and they support jumbo frames (MTU=9000), changing the MTU at the interface level works and I can see AoE packets (layer 2) bigger than 1500, it does reach +8000. RHCS supports changing the MTU in cluster.conf via the 'netmtu' option in the totem directive and theoretically it's supposed to double the throughput performance, quote from openais.conf man page:

> Increasing  the  MTU  from 1500 to 8982 doubles throughput performance from 30MB/sec to 60MB/sec as measured with evsbench with 175000 byte messages with the secauth directive set to off

I've changed it in cluster.conf and has been accepted by CCS but I can't see any IP (layer 3/4) packets larger than 1472, anyone had any luck changing that and worked for them?


  -- Abraham

Abraham Alawi

Unix/Linux Systems Administrator
Science IT
University of Auckland
e: a.alawi at auckland.ac.nz
p: +64-9-373 7599, ext#: 87572


More information about the Linux-cluster mailing list