On 6/14/07, Tom Haddon <tom@xxxxxxxxxxxxxxxxx> wrote:
On Thu, 2007-06-14 at 13:02 -0600, Serge Dubrouski wrote: > I wouldn't rely on balancer_manager for a such monitoring because I've > seen situations when it reported "Ok" status for workers that were > down. Is that a known bug?
I haven't seen any bugs reported for this in Bugzilla.
> > As alternative method you can use sniffing Apache error_log file for > the messages about disabled workers. You can OSSEC for that for > example. Is there any way of getting the current status (which workers are in and out) reliably, as this would surely rely on a good known state and then playing things forward from there? I've been seeing some messages in the error logs about workers being disabled, but when I test them they are plainly working fine, and there's no message about the workers being brought back into the loop.
Not that I know about. I monitor all downstream servers that I have configured in load balancer and load balancer itself using Nagios. It would be good to be able to make sure that Apache uses all those downstream server but I don't a simple way to do that. One way could be monitoring for incoming traffic/growing log files on those servers.
Thanks, Tom > > On 6/14/07, Tom Haddon <tom@xxxxxxxxxxxxxxxxx> wrote: > > Hi Folks, > > > > We've recently set up mod_proxy_balancer as a replacement for Pound, and > > I'm trying to work out the best way to monitor the load balancer to > > determine which servers are "in the loop" and which servers are "out of > > the loop". I'd ideally like to tie this in to Nagios so that I can get > > alerts if any of my servers gets taken out of the loop. > > > > At the moment I have the balancer-manager admin interface enabled for > > local connections only (i.e. from 127.0.0.1 on my apache server) and > > password protected, and have implemented something like the following > > for my Nagios check: > > > > command[check_beta_lb_1]=/usr/lib/nagios/plugins/check_http -H localhost > > -u /balancer-manager -a betaload:--password-- -s > > 'http://--dest-ip--:--dest-port--</a></td><td></td><td></td><td>1</td><td>Ok' > > > > This seems like a terrible hack, and I'm wondering if there's a better > > way to do this. > > > > Thanks, Tom > > -- > > Tom Haddon <tom@xxxxxxxxxxxxxxxxx> > > > > > > --------------------------------------------------------------------- > > 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@xxxxxxxxxxxxxxxx > > " from the digest: users-digest-unsubscribe@xxxxxxxxxxxxxxxx > > For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx > > > > > > --------------------------------------------------------------------- > 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@xxxxxxxxxxxxxxxx > " from the digest: users-digest-unsubscribe@xxxxxxxxxxxxxxxx > For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx -- ---------------------------------- Tom Haddon mailto:tom@xxxxxxxxxxxxxxxxx m +1.415.871.4180 www.greenleaftech.net --------------------------------------------------------------------- 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@xxxxxxxxxxxxxxxx " from the digest: users-digest-unsubscribe@xxxxxxxxxxxxxxxx For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx
--------------------------------------------------------------------- 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@xxxxxxxxxxxxxxxx " from the digest: users-digest-unsubscribe@xxxxxxxxxxxxxxxx For additional commands, e-mail: users-help@xxxxxxxxxxxxxxxx
|