Re: NFS client latencies

[Lee Revell <rlrevell@joe-job.com>]

On Wed, 2005-03-30 at 09:50 -0500, Trond Myklebust wrote:
> on den 30.03.2005 Klokka 09:26 (-0500) skreiv Lee Revell:
> > On Tue, 2005-03-29 at 18:18 -0500, Trond Myklebust wrote:
> > > ty den 29.03.2005 Klokka 18:04 (-0500) skreiv Lee Revell:
> > > > I am seeing long latencies in the NFS client code.  Attached is a ~1.9
> > > > ms latency trace.
> > > 
> > > What kind of workload are you using to produce these numbers?
> > > 
> > 
> > Here is the other long latency I am seeing in the NFS client.  I posted
> > this before, but did not cc: the correct people.
> > 
> > It looks like nfs_wait_on_requests is doing thousands of
> > radix_tree_gang_lookups while holding some lock.
> 
> That's normal and cannot be avoided: when writing, we have to look for
> the existence of old nfs_page requests. The reason is that if one does
> exist, we must either coalesce our new dirty area into it or if we
> can't, we must flush the old request out to the server.

But holding a spinlock for 3ms is not acceptable.  _Something_ has to be
done.  Can't the lock be dropped and reacquired after processing N
requests where N is some reasonable number?

Lee