Hi List.  This is a bit long because I'm not sure what's going on, need
a clue for where to look next, so I'm outlining our situation as well as
I can.

We have a two-server FOG running Windows Connector sessions for about
143 DTU's on Solaris 10/x86.  We applied Sun's latest Solaris patch set
and updated Sun Ray from 4.0 to 4.1 last Tuesday, but this morning was
the first the systems have been heavily used since then (lots of
students coming back from break).

I got word this morning of a lot of DTU's showing hung sessions.  No
idea what caused it but a warm restart of the servers appeared to get
things unstuck, so I'd chalk it up to gremlins and move on, but now 
there's some other odd stuff going on.

Normally, each machine in our FOG will get roughly half the sessions,
but currently the head node has 130, and the other only has 13.  Since
the trouble we had this morning I've been keeping the web GUI open and
watching sessions come and go, and that count is staying fairly
constant.

We've continued to see sessions getting stuck in "Idle" state -- with a
surge of them since I started writing this message -- but they're only
getting stuck on the head server, not the secondary.  I'm starting to
suspect FOG communication issues.

"utsession -p" doesn't show the sessions with state "I", but does show
them with missing Unix ID's, "???" where there would normally be
"utku##".  Mentioning that in case it's a clue to telling us what's
going on.

Looking through /var/opt/SUNWut/log/messages, I don't find anything that
jumps out as an error, but that's a busy file so I could easily be
overlooking something.  I'm sure I could also be looking at the wrong
log file.

Any hints?

Thanks for reading,
--Michaael

_______________________________________________
SunRay-Users mailing list
[email protected]
http://www.filibeto.org/mailman/listinfo/sunray-users

Reply via email to