Well, nfsstat I can do, and this is the pertinent section. Is this high for getattrs and readdirplus? Version 3: (323847 calls) null getattr setattr lookup access readlink 7 0% 177796 54% 1570 0% 2049 0% 37509 11% 5 0% read write create mkdir symlink mknod 32004 9% 47471 14% 351 0% 221 0% 5 0% 0 0% remove rmdir rename link readdir readdirplus 892 0% 49 0% 0 0% 0 0% 0 0% 22931 7% fsstat fsinfo pathconf commit 619 0% 16 0% 0 0% 352 0%
On 5/4/06, Matty <matty91 at gmail.com> wrote: > > On Thu, 4 May 2006, Joe Little wrote: > > > I'll try the nocto option. Our average file size (I believe) is in the > > low K's -- perhaps between 1K-8K. An example where this gets really > > slow is a prof with untarred versions of every linux kernel release > > for source code analysis. It processes these directories about 2K/sec. > > I will confess to not knowing how to measure latency. I haven't > > observed the issue with UFS, but I haven't been trying much since UFS > > doesn't interest me as much. That said, our UFS only Solaris 9/10 > > machines haven't had this issue. > > I have found tcptrace, snoop -vv, and the DTrae profile and io providers > invaluable for putting together data flow diagrams with latency > measurements. There is also some useful information in nfsstat, which > might help you better understand activity on the client and server. > > Hope this helps, > - Ryan > -- > UNIX Administrator > http://daemons.net/~matty > _______________________________________________ > nfs-discuss mailing list > nfs-discuss at opensolaris.org >
