Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cornell!uw-beaver!mit-eddie!rutgers!att!ulysses!andante!alice!debra
From: debra@alice.UUCP (Paul De Bra)
Newsgroups: comp.unix.xenix
Subject: Re: apparent cron lockup
Message-ID: <8870@alice.UUCP>
Date: 5 Feb 89 00:47:12 GMT
References: <286@berner.UUCP> <611@dms.UUCP> <13073@steinmetz.ge.com> <455@berner.UUCP>
Reply-To: debra@alice.UUCP ()
Organization: AT&T, Bell Labs
Lines: 30

In article <455@berner.UUCP> richard@berner.UUCP (Richard Greenall) writes:
>In article <13073@steinmetz.ge.com> davidsen@crdos1.UUCP (bill davidsen) writes:
>> [ doubtful explanation of cron problem deleted ]
>
>I don't think that that was the entire problem with cron.  I also
>run uucp a couple times an hour and still get the cron lockup
>quite regularly with 2.3.1 I even added the /bin/true to run about
>every half hour with little success.  I even went back to the
>cron in version 2.2.1 and installed that, and STILL got the
>same symptoms.

The cron problem exists in other versions of Unix as well. The time between
events is irrelevant since cron is supposed to wake up every minute to check
whether the crontab (or crontabs in newer releases) has changed.

The only instances of cron falling asleep that I have seen occured in a
system where cron had been swapped out. Apparently there is a bug in some
Unix systems (not just Xenix and not only on intel processors!) that causes
processes that are swapped out while sleeping to remain asleep for ever. 
(I mean waiting for the end of sleep(), not waiting for some other event
like I/O).

I never noticed the problem in my Xenix box because it has enough memory
to avoid all swapping.

Paul.
-- 
------------------------------------------------------
|debra@research.att.com   | uunet!research!debra     |
------------------------------------------------------