Michael,
Had forgotten logging into B queue altogether! At least that would enable
looking for culprit. We don't do any playing with queues - throughput is 99%
adequate without that. Only a few of us have access to mpe prompt, and most
don't know about those things.
thanks, will file this just in case.
john
-----Original Message-----
From: Michael Anderson [mailto:[log in to unmask]]
Sent: Thursday, 29 March 2007 3:55 AM
To: [log in to unmask]; john pitman
Subject: Re: [HP3000-L] strange hang....try 2
Memory Dump? What were the scheduling sub-queues set to?
Check for any users running in the 'B' or 'A' Que.
I've had this problem before where the system would seem to hang.
Could be someone playing with the (TUNE) sub-queues or someone run an
application in the 'A' or 'B' queues, locking a system resource or two.
If it happens again:
My Solution is to logon to the 'B' queues, HELLO
MICHAEL,MANAGER.SYS;PRI=BS
Some M.P.E. experts may disagree with this , arguing that "You never
log a session on to the 'B' queues", and they are Correct!
My point is that one of the few times to log a session on the 'B'
queues is in the exact scenario you just described, you ever mentioned
re-boot.
;PRI=BS; You'll be logged on instantly if their is nothing wrong with
your machine,
and you can run something like GLANCE or SHOWPROC, (nothing that will
hog resources) to find out what is bringing your machine to it knees.
Your solution worked good too.
Later,
Mike.
>>> john pitman <[log in to unmask]> 3/27/2007 11:43 pm >>>
Monday this week, late in the day, system seemed to hang. I tried to
start a
fresh session, got through login, no prompt. Within minutes others
reported
this. Some users inside running programs still seemed to be able to
work for
a while . no prompt on the console, but messages were appearing there,
but
could not get a prompt - not network specific. Some attempts at login
then
got 'cant initiate new sessions now' msg. Using hipri forced a login,
but no
prompt.
After a while we gave up and rebooted it, and its been up ok since.
Anybody have any idea what could tie it up so tightly that no ci could
get
started? There was activity on the disks that sort of looked normal.
N series, 1GB RAM, 2 HDD, 7gb free space, 1 x 220Mhz cpu. Max sessions
280,
have seen it pass 250 a few times.
thanks,
john
* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
|
