I have initally sent this mail to port-mac68k. 'The' ps problem explained
hereafter seems to be confirmed on macintosh but pehaps does not confine
itself to this platform. It would be interesting to know if this bug can
be reproduced on other implementations before starting to track it down.

>It appears that the 'ps' command can not manage under heavy load conditions.
>I have noticed it during a 'daily' cron job: The 'ps aux' output starts to
>print
>slowly and just hang. I have to kill this unfortunate process to end its
>miserable life.
>When a shell script executes, a lot of processes start and die at high rate.
>The 'ps' command probably makes its way through kernel lists and arrays while
>they are changing quickly. Yet, I presume that 'ps' is more than aware
>about it,
>and manipulates kernel data with appropriate care...
>
>Well, let us start to confirm the problem before digging the code out:
>Did anyone noticed such 'ps' failure ?

Francois Pays