Using Solaris 10 on Sparc, how's the best way to hunt down NFS
performance issues if you suspect a single client of hogging the NFS
resources on a home directory server using ZFS for the NFS home
shares?

I know there are some Dtrace NFS tools along with netstat, nfsstat,
and snoop. None of which I use everyday.

Does anyone one a "best practice" or procedure they use all the time
to track this type of issue?

I was thinking a snoop -d ce0 |egrep "NFS R" and basically seeing
which client is doing a massive amount of called to NFS on that
ethernet port at the time the nfsd is spiking the CPU times.

This is something I don't normally have to troubleshoot.

Any advice is welcome.

Thanks,