Wow. Holy Mackerel!
Well, we do run nagios, though I avoid it like the plague. But perhaps
I can grok from this some ideas how to make my own novel tests from
within Monit.
Thanks. :)
On 7/14/14 12:10 PM, Michelle Sullivan wrote:
Paul Theodoropoulos wrote:
I'm currently monitoring bucardo using Monit - but only the most
basic, checking PID. I've found on a couple of occasions that bucardo
will have hit a wonky streak, and stopped running a sync, which the
simple PID checking is incapable of recognizing as an issue.
Is anyone else monitoring bucardo with Monit?
I am not.
Do you have a novel way of checking to see if all is awesome in
bucardo-land, and alerting if not?
Define novel ;-) ... Here's my Nagios check function:
http://flashback.sorbs.net/packages/check_bucardo_repl.pl.txt
The simplest way to do it is add a table 'bucardocheck' to the sync you
want to monitor (table name is definable) with columns of pk (INT) and
ts (TIMESTAMPZ) than then define the command in nagios with the relevant
params.. (see help).. if you're doing multimaster set a different pk
row for each master or you'll get continual conflict checking (which is
ok, but it slows stuff down and screws with the lag calculation)
Regards,
Michelle
Here's my dirt-simple monit config as an example --
check process bucardo with pidfile /var/run/bucardo/bucardo.mcp.pid
start program = "/etc/init.d/bucardo start"
stop program = "/etc/init.d/bucardo stop"
alert [email protected] on { pid exec nonexist status } with
reminder on 5 cycle
if 3 restarts within 4 cycles then alert
--
Paul Theodoropoulos
www.anastrophe.com
------------------------------------------------------------------------
_______________________________________________
Bucardo-general mailing list
[email protected]
https://mail.endcrypt.com/mailman/listinfo/bucardo-general
--
Paul Theodoropoulos
www.anastrophe.com
_______________________________________________
Bucardo-general mailing list
[email protected]
https://mail.endcrypt.com/mailman/listinfo/bucardo-general
