Earlier today, Darrell Shively wrote:
> We have htdig on our web server, and it works very well indeed.
> Manual indexing, etc. worked great and everything seems fine - with one
> tiny exception: the cron task (in /etc/cron.daily/htdig-dbgen) sends me
> the following email every day:
>
> > Subject: Cron <root@squid> run-parts /etc/cron.daily
> > From: "CronDaemon" <[EMAIL PROTECTED]>
> > To: root
> >
> > htmerge: Word sort failed
Then Denis Barbier wrote:
> > on a Linux box, i had to replace the call to htmerge by
> > /bin/tcsh -c /usr/sbin/htmerge
> >
> > A bug in bash ?
To which I responded:
> Hi, guys! I guess I wasn't thorough enough in testing this under Red
> Hat 5.0 and 5.1. It works fine under 4.2. I'm trying to track down
> the problem right now, but not having a lot of luck so far. For some
> reason, pclose() is returning an ECHILD error after the sort, which
> suggests that it's giving the PID of a non-existant process to wait4().
> A bug in glibc's pclose(), perhaps?
>
> I don't know why running htmerge from tcsh fixes it. So does running
> it from /bin/ash. Changing the first line of rundig to "#! /bin/ash"
> also fixes it. This is very bizarre. I'd sure like to get to the
> bottom of it. I don't see why starting htmerge from bash causes a
> wait4() for htmerge's child process to fail, but starting it from
> ash or tcsh doesn't. Anyone have any ideas about this?
OK, I've done some more digging, and here's what I discovered. I replaced
the glibc version of htdig-3.1.0b1 with the one I compiled on RH 4.2,
and even got cron to run the rundig script using the non-glibc bash from
4.2, and I still got the "Word sort failed" error! So, I got rid of
vixie-cron, on my 5.1 system, and installed vixie-cron-3.0.1-19.i386.rpm
from RHL 4.2, then ran rundig from cron (using the glibc versions of
bash & htdig), and it worked without the error!
It's way too late to look into this further tonight, so maybe next week
I'll take a closer look at what's changed in crond to cause this. Maybe
a glibc bug affecting it? Maybe a patch that wasn't well thought out?
I'll let you know if I find anything, but for now I'm sticking to the
/usr/sbin/crond binary from 4.2.
I also found an interesting message on the htdig mailing list archives
on www.htdig.org. On July 10th, Rob Willett reported similar findings
to my earlier one (I should have checked the archives first), and also
said his tests seem to indicate it's time-sensitive. (This was with
htdig 3.0.8b2, I believe, on RH 5.1.)
--
Gilles R. Detillieux E-mail: <[EMAIL PROTECTED]>
Spinal Cord Research Centre WWW: http://www.scrc.umanitoba.ca/~grdetil
Dept. Physiology, U. of Manitoba Phone: (204)789-3766
Winnipeg, MB R3E 3J7 (Canada) Fax: (204)789-3930
----------------------------------------------------------------------
To unsubscribe from the htdig mailing list, send a message to
[EMAIL PROTECTED] containing the single word "unsubscribe" in
the body of the message.
