[Beowulf] NFS Performance using SLES 10

[Mark Hahn]

> I have a small two-node cluster with AMD64 Opteron processors.  It was 
> running Centos 4 and NFS performance was good while not hindering 
> performance.  I replaced Centos with SLES 10 about 2 weeks ago and just now 
> got around to really hitting it hard submitting jobs.  Processes that 
> normally took about 15 seconds such as normal IO (each processor writes its 
> own data file) now require 10-15 minutes.  When I observed the IO status, I

the magnitude of this difference is not just a matter of tuning.
for instance, NFS exports have changed over the past few years 
in whether they require the 'async' flag.  that's something that can 
easily cause big differences in performance.  do you have a reasonable
number of nfsd's running?  if UDP, are you sure you're not seeing some 
problem with fragmented packets?