UNIX for Advanced & Expert Users

Hi all,

we have been noticed that few of processes failed with no reason.
when I chcked the log messages, I have got the following error messages writen to ktlog
just a few minutes before!

When the kernel writes such warning meassages?
will the system sends/generates any signals like
SIF_TERM, as we have naticed on one of our processess(handled for signals) failed because of this signal?
what should I do?


thanks
Bhupal

3bdd7f76 11:10:30 tolog/nocons/console - *TQ* WARNING: <2.88> himalya01 Long Run Queue
3bdd7f76 11:10:30 tolog/nocons/console - The CPU is busy.
3bdd7f76 11:10:30 tolog/nocons/console - Some processes may need to be terminated.
3bdd7f76 11:10:30 tolog/nocons/console - Additional CPUs may be needed.

Bhupal,
SIGTERM or signal-15 is a graceful way of terminating a process. this signal can also be generated the user by using "Kill" or may even be generated in your case by the kernel. That is what I can presume after seeing the log files.

I cant tell much about the actual reason for termination of the process by seeing these three lines of your file. But may be ur CPU is too busy that is what I see in logs. but in case, it should generate a signal

SIGXCPU CPU time limit exceeded.

Add a singal handler in your process for this signal and that can help you find out whether this isthe actual problem.

Good luck