From: Gavin Kelman (gavin@metahusky.net)
Date: Wed Nov 06 2002 - 11:16:55 EST
Our V880 with Solaris 9 and the latest patch cluster installed, mounts
our Solaris 8 file server. The Solaris 8 box is probably about 3 months
behind the latest patch cluster (always on, planned maintenance is rare).
/srsdata sits on a t3 attached to the solaris 8 box.
The Solaris 9 box has been up and running for about 4 days now mounting
/srsdata on the Solaris 8 box. Today, people on other machines (we have
Irix, tru64, hp-ux, linux, solaris 7 & 8) complain that /srsdata is
impossibly slow to write to. Reading is fine, writing is so slow, touching
a file there takes about 45 seconds. The same is true if done locally on
the fileserver itself, avoiding NFS.
I spent a good 2 hours watching NFSOPS, 800-1000 at this time of day,
totally normal. I checked syslog, dmesg, and vxdiskadm for errors on failed
disks or similar, and nothing seemed amiss. In desperation, I unmounted
/srsdata from the solaris 9 box, and suddenly everything returns to normal.
/srsdata becomes fast to write to again.
What's going on? Why is our Solaris 9 box effectively Denial Of Servicing
our Solaris 8 file server? I'm about to get in touch with Sun about this,
but wondered if anyone had any ideas what's going on....
Cheers,
Gavin.
-- Gavin Kelman Phone : 07967 360 822 Systems Administrator Email : gavin@metahusky.net _______________________________________________ sunmanagers mailing list sunmanagers@sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers
This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:25:14 EDT