> -----Original Message-----
> From: Phil Yantis
> John Penny asked to see the actual stdlist. Here it is:
> >>L ALL
>       2   JOB MSJOB,MGR.MINISOFT,MM.
>       3   Priority = DS; Inpri = 8; Time = 6000 seconds.
>       4   Job number = #j13636.
>       5   WED, JAN 16, 2002,  9:50 AM.
[...]
>      52   :SETVAR MSRUNPRI "CS"
>      53   :SETVAR MSSERVER000002 "30001 0 ...
>      54   :SETVAR MSSERVER000003 "30003 0 ...
>      55   :SETVAR MSSERVER000004 "30006 0 ...
>      56   :SETVAR MSSERVER000008 "30002 0 ...
>      57   :SETVAR MSSERVER000016 "30004 0 FMSRVR.MM.MINISOFT;PRI=DS S"
>      58   :RUN MSSERVER.MM.MINISOFT;INFO="30000 SERVER.MM.MINISOFT"
>      59
>     246   JOB exceeds time limit. (CIERR 6028)
>     247   CPU sec. = 33.  elapsed min. = 44405.  SAT, FEB 16,
> 2002,  5:55

hmmm... with 44,000+ connect minutes, only 33 seconds of actual CPU time
seems rather low for a "server" type process.  I seem to recall, however,
that when you have son processes, the reported CPU time is that of the
topmost process (i.e., the one directly executed from the CI; the CPU time
of a child process may not be reported...)

I think you can see this for yourself by running a CPU intensive program,
pressing BREAK during the run, then SHOWME within break will show the CPU
time of the parent CI process but not your current process -- that doesn't
get "folded in" until the process actually quits.  I'm thinking that if the
system is terminating the job, the "reported" time may not be updated by the
time the job actually terminates (dunno why not, but I seem to recall seeing
this during my days at VEsoft...)

Tom

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *