On Thursday 21 May 2009 01.29.07 you wrote:
> Hi gang...
>
> It would be great if people running NFS (soft, and hard mounts)
> could give this diff a spin. what it does essentially is to redo
> the retransmission booking by keeping count of maximum rexmits in
> one place (the nfsmount structure) and kill deadthresh, (an nqnfs
> left over); that causes a slight behaviour change when printing
> the "not responding" message for dead servers (it does it now
> a tad bit sooner then before (at 4 rexmits, not 9 iirc)).
>
> this diff is first of a few to get the stale/dead nfs mount point
> issued fixed.

$ mount |fgrep nfs
<host>:/mnt/data/music on /mnt/serenity_music type nfs (nodev, nosuid,
read-only, v3, udp, soft, intr, timeo=100)

Working OK with it. Should I do something with it (plug the network out, kill
the nfsd), to further test it?

Daniel

--
LIVAI Daniel
PGP key ID = 0x4AC0A4B1
Key fingerprint = D037 03B9 C12D D338 4412  2D83 1373 917A 4AC0 A4B1

Reply via email to