Re: monit and ganglia

[Andrew Ring]

Thank you that works great.


For the archives:
For my CentOS 6.3 system, I had to install monit 5.6 (5.1.1-4 is not 
sufficient).
Other limitations were imposed, but this stripped down script worked.

# gmetad
check host gmond with address localhost
  start program = "/etc/init.d/gmond start"
  stop program  = "/etc/init.d/gmond stop"
  if failed host localhost port 8649 expect "DOCTYPE GANGLIA_XML"
    then restart
  if 5 restarts within 5 cycles then alert




On 01/14/2014 10:42 PM, Daniel Rich wrote:
> Here's what I use. It's on FreeBSD, so you'll need to change the paths 
> if you're running on Linux.
>
>     #
>     # Ganglia
>     #
>
>     # gmetad
>     check process gmetad matching "sbin/gmetad"
>       start program = "/usr/local/etc/rc.d/gmetad start"
>       stop program  = "/usr/local/etc/rc.d/gmetad stop"
>       if failed host localhost port 8651 expect "DOCTYPE GANGLIA_XML"
>         then restart
>       if 3 restarts within 5 cycles then timeout
>       group server
>
>     # gmetad
>     check process gmond with pidfile /var/run/gmond.pid
>       start program = "/usr/local/etc/rc.d/gmond start"
>       stop program  = "/usr/local/etc/rc.d/gmond stop"
>       if failed host localhost port 8649 expect "DOCTYPE GANGLIA_XML"
>         then restart
>       if 3 restarts within 5 cycles then timeout
>       group server
>
>
> On 1/14/14, 10:53, Andrew Ring wrote:
> > Hello All,
> >
> > Does anyone have a working script for configuring Monit to monitor 
> > Ganglia's gmond daemon?
> >
> > When gmond is running properly, manually testing with "telnet 
> > localhost 8649" will return some XML and exit.  When gmond is off or 
> > has crashed, the results is "Connection refused".
> >
> > With Monit I have tried "if failed host localhost port 8649 then 
> > restart", but this fails to recognize when gmond is running.
> >
> > Thank you,
> > Andrew
> >
> >
> > --
> > To unsubscribe:
> > https://lists.nongnu.org/mailman/listinfo/monit-general
>
> --
> Dan Rich<address@hidden>  |http://www.employees.org/~drich/
>                                 |  "Step up to red alert!"  "Are you sure, sir?
>                                 |   It means changing the bulb in the sign..."
>                                 |          - Red Dwarf (BBC)
>
>
> --
> To unsubscribe:
> https://lists.nongnu.org/mailman/listinfo/monit-general

--
Andrew Ring
HPC Systems Administrator
University of Minnesota Twin Cities
Phone: 612-624-8860                 Fax: 612-624-8861
Mail:                               Office:
599 Walter Library                  543 Walter Library
117 Pleasant St S.E.                117 Pleasant St S.E.
Minneapolis, MN 55455               Minneapolis, MN 55455