Comment 23 for bug 585657

Revision history for this message
Sean Clarke (sean-clarke) wrote :

I am hitting this problem under 10.10 x64:

uname -a
Linux enterprise 2.6.35-27-server #47-Ubuntu SMP Fri Feb 11 23:09:19 UTC 2011 x86_64 GNU/Linux

I've reported it under a couple of other open bugs relating to issues around this area, can you let me know what kernel version to expect the change to be rolled out in? It is a huge problem for us as we run KVM images over NFS and this happens every time.

After it happens, we also get "false" timeouts on the NFS server and the whole system stutters and stalls. The NFS server continues to serve files to other systems and can be ping'd from the failed client, it even serves files to it - but you get very regular (5 seconds?) timeout messages in the logs:

[15594.126931] nfs: server XXXXXX not responding, timed out
[15598.336861] nfs: server XXXXXX not responding, timed out
[15602.546851] nfs: server XXXXXX not responding, timed out
[15606.757764] nfs: server XXXXXX not responding, timed out
[15610.966788] nfs: server XXXXXX not responding, timed out
[15615.176756] nfs: server XXXXXX not responding, timed out

PING XXXXXX 56(84) bytes of data.
64 bytes from XXXXXX: icmp_req=1 ttl=64 time=0.097 ms
64 bytes from XXXXXX: icmp_req=2 ttl=64 time=0.059 ms
64 bytes from XXXXXX: icmp_req=3 ttl=64 time=0.079 ms