To whoever administers this: - can ping www.wlug.org.nz - http to www.wlug.org.nz, www2.wlug.org.nz, wlug.org.nz times out so suspect web server hanging or similar - coming in from Xtra DSL in case network issue
On Wed, Feb 23, 2005 at 08:23:37AM +1300, Ian McDonald wrote:
To whoever administers this: - can ping www.wlug.org.nz - http to www.wlug.org.nz, www2.wlug.org.nz, wlug.org.nz times out so suspect web server hanging or similar - coming in from Xtra DSL in case network issue
Yes, for some reason, the apache webserver on the server occasionally fails to restart gracefully after doing the daily logrotate in the mornings. We thought we had fixed it, but it seems to still be occuring sometimes. John
John R. McPherson wrote:
On Wed, Feb 23, 2005 at 08:23:37AM +1300, Ian McDonald wrote:
To whoever administers this: - can ping www.wlug.org.nz - http to www.wlug.org.nz, www2.wlug.org.nz, wlug.org.nz times out so suspect web server hanging or similar - coming in from Xtra DSL in case network issue
Yes, for some reason, the apache webserver on the server occasionally fails to restart gracefully after doing the daily logrotate in the mornings. We thought we had fixed it, but it seems to still be occuring sometimes.
John
When I was working at TKI (www.tki.org.nz), we had the exact issue. Are you performing a restart ? We found we HAD to stop then start, otherwise it would hang every night after the logrotation. My 2 cents. - Drew
On Wed, Feb 23, 2005 at 09:40:02AM +1300, Drew Broadley wrote:
John R. McPherson wrote:
Yes, for some reason, the apache webserver on the server occasionally fails to restart gracefully after doing the daily logrotate in the mornings. We thought we had fixed it, but it seems to still be occuring sometimes.
When I was working at TKI (www.tki.org.nz), we had the exact issue. Are you performing a restart ? We found we HAD to stop then start, otherwise it would hang every night after the logrotation.
Yes, we had this fixed before we rebuilt the server last month. It's due to logrotate doing a 'reload', which doesn't seem to work properly. I'm just trying to figure out why apache is getting a reload every morning when logrotate should only be doing apache's log file weekly. It looks like the workarounds are to do a stop-start/force-reload instead of a graceful reload, or to use a pipe and cronolog for logging instead of logrotate... John
It looks like the workarounds are to do a stop-start/force-reload instead of a graceful reload, or to use a pipe and cronolog for logging instead of logrotate...
We ran into the same problem; restart/reload doesn't seem to close the log files. stop sometimes takes a while to happen so you want a short pause before starting again. Just to be sure I added a second restart 25 seconds later in our postrotate script..
On Wed, Feb 23, 2005 at 08:23:37AM +1300, Ian McDonald wrote:
To whoever administers this: - can ping www.wlug.org.nz - http to www.wlug.org.nz, www2.wlug.org.nz, wlug.org.nz times out so suspect web server hanging or similar - coming in from Xtra DSL in case network issue
Yes, for some reason, the apache webserver on the server occasionally fails to restart gracefully after doing the daily logrotate in the mornings. We thought we had fixed it, but it seems to still be occuring sometimes.
John _______________________________________________ wlug mailing list | wlug(a)list.waikato.ac.nz Unsubscribe: http://list.waikato.ac.nz/mailman/listinfo/wlug -- Oliver Jones » Roving Code Warrior
I've had this same problem on a RHEL 2.1 server I maintain. It is something to do with a re-entrancy in PHP (or something). Something about the signal handling in PHP causes it to segfault on the second SIGHUP. So basically it bails every second day. I altered the log rotation scripts to do a service restart instead. Solved the problem. Not ideal but it works. Regards On Wed, 2005-02-23 at 08:40 +1300, John R. McPherson wrote: oliver(a)deeperdesign.com » +64 (21) 41 2238 » www.deeperdesign.com
participants (5)
-
Drew Broadley -
Ian McDonald -
John R. McPherson -
Oliver Jones -
zcat