httpd-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jez Hancock <jez.hanc...@munk.nu>
Subject Re: [users@httpd] [httpd] Undescribeble problems, but I'll try...
Date Mon, 15 Dec 2003 23:57:03 GMT
On Fri, Dec 12, 2003 at 02:05:56PM +0100, debian@elastic.nl wrote:
> I have this really weird Apache Httpd problem
Aren't they all? :)

> When xx weeks/days have past, Apache just stops servering webpages. The
> daemon is still running and all seems OK. System load is normal and all
> other services just work perfectly.
> 
> The daemon can not be restarted normally and a graceful restart will result
> in this:
> 
>     /var/www/bin/apachectl graceful: httpd not running, trying to start
>     /var/www/bin/apachectl graceful: httpd started
> 
>     [Wed Dec 10 11:44:46 2003] [crit] (98)Address already in use: make_sock:
>     could not bind to port 80
> 
> Seems like apache keeps on logging for a while, since some httpd request are
> more than once in de logs at the time this weird problem occurs. Could be
> because clients keep reloading the page Apache won't serve. Since this is
> not a very busy webserver, logging either stops or visitors go away... I
> am not sure.
> 
> I stop the daemon with a "kill -HUP" and then start it again... all is fine
> then.
Note that a hangup/HUP signal is generally synonymous with a 'restart' -
generally if someone says 'I HUP'd the server' they mean they sent a
HUP/hangup signal to it to restart or 'bounce' it.  With this in mind
given your description of the problem above this is probably not what
you want to do.

More effective would be to send a KILL signal (signal 9) - for more info
on kill see your manpage for kill.

(also note though that how a process handles a given signal depends on
how the underlying code is written).

<snip>
> Apache 1.3.28 was already replaced by 1.3.29 (not a Debian package, but
> compiled from source). The motherbord and cpu is alse changed, but not
> because of this problem.
Perhaps try the debian package instead?

> Nothing is in the Crontab that could generate this error, I already checked.
> 
> After the accident I put server-status on and I'm logging its output hoping
> to see something when the problem happens again. Right now the server-status
> does not show any thing suspicious (see below this mail).
I was going to suggest you could add a 5min cronjob to test the status
of the server and if there's no response email you.

Given it's not a busy server though perhaps it might be best to just try
reinstalling from scratch (with the debian package)?

-- 
Jez Hancock
 - System Administrator / PHP Developer

http://munk.nu/
http://jez.hancock-family.com/  - personal weblog
http://ipfwstats.sf.net/        - ipfw peruser traffic logging

---------------------------------------------------------------------
The official User-To-User support forum of the Apache HTTP Server Project.
See <URL:http://httpd.apache.org/userslist.html> for more info.
To unsubscribe, e-mail: users-unsubscribe@httpd.apache.org
   "   from the digest: users-digest-unsubscribe@httpd.apache.org
For additional commands, e-mail: users-help@httpd.apache.org


Mime
View raw message