As promised, further information about crons that (at random) just die
and dump core. I mentioned in my previous summary that I hadn't actually
managed to track down the relevant sun patch yet, and lo and behold, I
got a message within the day from
David Moline <firstname.lastname@example.org>
(to whom I'm very grateful), mentioning the following (Solbourne) patch:
> Patch applies to:
> 4.1A.1 release
> Affected bugs:
> Patch name:
> P911121001: cron dies on correct entries
> This patch is the equivalent of Sun patch 100402-01.
> The ``usr'' structure of an at/cron job could be removed while the job
> was still running due to a race condition. This causes a core dump at
> the end of the job, when cron attempts to print out the now-missing
> The symptom is the routine ``logit'' in the stack backtrace from the
> dead cron.
I've grabbed the Sun patch mentioned (archie listed a whole bunch of places
- I'll mail anyone who wants the list), and this should be the last you hear
from me on this subject unless it fails to fix the problem (touch wood).
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:06:39 CDT