[Gluster-devel] 3.5.1qa4 performances

[Emmanuel Dreyfus]

On Tue, Dec 24, 2013 at 02:57:51AM -0500, Pranith Kumar Karampuri wrote:
> > And what about the looping ops getting Stale NFS file handle?
> Could you elaborate please? I did not understand the question

glustershd.log and brick log on two machines exhibit a lot of 
message per second about LOOKUP and GETXATTR getting  Stale NFS file handle
error. I posted some example in my first message.

It loops on the same files, therefore I suspect something wrong in
healing process.

-- 
Emmanuel Dreyfus
manu at netbsd.org