PDA

View Full Version : Problem with update to apache2-2.2.3-16.42.2.x86_64/sles 10



berndgsflinux
20-Dec-2011, 21:26
Hi,

i have three sles 10 sp4 boxes with apache, and i'm having problems. On
thursday morning, all three boxes installed apache2-2.2.3-16.42.2.x86_64
and the relating packages. This is done automatically via a cron job.

On thursday and friday on all three boxes ran jobs in /etc/cron.daily
(on different times of course). Skripts in this directory are e.g.
logrotate.
On all three boxes logrotate created new accesslogs for apache, and
restarted. Directly afterwards my monitoring-systems starts screaming
that my webservers are not available any longer. Unfortunally i've been
for two days off, in particular thursday and friday. My collegues (they
are scientists and don't know much about systemadministration) didn't
know what to do. They didn't check if httpd was running. They waited for
a time, afterwards they decided to reboot. After reboot everything was
fine.
If i'm doing now a manual restart via /etc/init.d/apache2 reload,
everything is going fine.

Is anyone else experiencing problems with the above mentioned apache
update ?


Bernd


--
berndgsflinux
------------------------------------------------------------------------
berndgsflinux's Profile: http://forums.novell.com/member.php?userid=5742
View this thread: http://forums.novell.com/showthread.php?t=449855

MoserHans
21-Dec-2011, 17:56
berndgsflinux;2162714 Wrote:
> i have three sles 10 sp4 boxes with apache, and i'm having problems. On
> thursday morning, all three boxes installed apache2-2.2.3-16.42.2.x86_64
> and the relating packages. This is done automatically via a cron job.
>
> On thursday and friday on all three boxes ran jobs in /etc/cron.daily
> (on different times of course).
This is not "of course". Go to /etc/sysconfig/cron and set DAILY_TIME
to whatever time you like - on all servers accordingly. :)
Now cron.daily is running on all server at one time.

> Skripts in this directory are e.g. logrotate.
> On all three boxes logrotate created new accesslogs for apache, and
> restarted. Directly afterwards my monitoring-systems starts screaming
> that my webservers are not available any longer. Unfortunally i've been
> for two days off, in particular thursday and friday. My collegues (they
> are scientists and don't know much about systemadministration) didn't
> know what to do. They didn't check if httpd was running. They waited for
> a time, afterwards they decided to reboot. After reboot everything was
> fine.
> If i'm doing now a manual restart via /etc/init.d/apache2 reload,
> everything is going fine.
>
> Is anyone else experiencing problems with the above mentioned apache
> update ?
These are logs for. Look at them and find out, if they can tell you
anything that happened. :)


--
MoserHans
------------------------------------------------------------------------
MoserHans's Profile: http://forums.novell.com/member.php?userid=53101
View this thread: http://forums.novell.com/showthread.php?t=449855

berndgsflinux
22-Dec-2011, 13:46
Hi Moserhans,

thank you for your very helpful post.

MoserHans;2163058 Wrote:
> This is not "of course". Go to /etc/sysconfig/cron and set DAILY_TIME to
> whatever time you like - on all servers accordingly. :)
> Now cron.daily is running on all server at one time.
This is not what i've been asking for.

MoserHans;2163058 Wrote:
> These are logs for. Look at them and find out, if they can tell you
> anything that happened. :)
Oh, i never heard of logs. What are logs, what is their purpose, and
where can i find these misterious logs ?


--
berndgsflinux
------------------------------------------------------------------------
berndgsflinux's Profile: http://forums.novell.com/member.php?userid=5742
View this thread: http://forums.novell.com/showthread.php?t=449855

MoserHans
23-Dec-2011, 09:36
berndgsflinux;2163230 Wrote:
> thank you for your very helpful post.
This, as the rest of your post, is irony, I get it.

About the logs:
In most cases it is difficult enough to tell "what happened" by looking
at what is in the logs.
You did not provide anything like that. The fact that nobody else
responded to your post may have to do with that.
Did you check what your logrotate scripts should do and what they have
really done (logs)?


--
MoserHans
------------------------------------------------------------------------
MoserHans's Profile: http://forums.novell.com/member.php?userid=53101
View this thread: http://forums.novell.com/showthread.php?t=449855

berndgsflinux
23-Dec-2011, 15:26
Hi,

following the logs, everything went fine. The cronjobs in
/etc/cron.daily reported via email to me:

===========================
running daily cronjob scripts

SCRIPT: output (stdout && stderr) follows

Reload httpd2 (graceful restart)..done
SCRIPT: logrotate
------- END OF OUTPUT


SCRIPT: output (stdout && stderr) follows

quotacheck: Scanning /dev/sda2 [/] quotacheck: Old group file not
found. Usage will not be substracted.
quotacheck: Checked 15000 directories and 374947 files
SCRIPT: quota.sh
------- END OF OUTPUT

===========================

Following to this, everything went fine. The log reported just about
backup-scripts for databases running fine during this time.
I think the graceful restart of httpd after logrotate rotated some
log-files from apache caused the downtime, because my monitoring-system
started screaming directly afterwards.
Although the email says: Reload httpd2 (graceful restart)..done
I cloned the regarding server from the backup in the state directly
before applying the apache update.
I will now run the update, then do a "/etc/init.d/apache reload" and
see what happens.

I will keep you informed.

Merry christmas and a happy new year to everyone in this forum.



Bernd


--
berndgsflinux
------------------------------------------------------------------------
berndgsflinux's Profile: http://forums.novell.com/member.php?userid=5742
View this thread: http://forums.novell.com/showthread.php?t=449855