Re: cron style service poll time not functioning since update to 5.10

[Paul Theodoropoulos]

On 11/17/14 2:29 PM, Jan-Henrik Haukeland wrote:
> > On 17 Nov 2014, at 19:28, Paul Theodoropoulos <address@hidden> wrote:
> >
> > I have a service check that I disable while backups are running on a particular 
> > server. Since upgrading from 5.8.1 to 5.10, the cron declaration appears to no 
> > longer work - I've started getting alerts during that time since the upgrade.
> >
> > Here's how I have it set up -
> >
> > check program servercomm-db1 with path /usr/local/bin/check_server_wrapper with 
> > timeout 25 seconds
> > not every "40-50 0 * * *"
> >
> > I checked the manual, and that appears to be a valid config. Thoughts?
> This should work, please check that you do not have a a second (old) monit 
> instance running in the background or that the every-cron string is specified 
> as you think it is. Right now, Monit is specified to skip this check every 
> night between 00:40-00:50.
Thanks. The above declarations were cut & paste directly from the conf 
file, and I've confirmed that it is indeed running 5.10, with no other 
monit executables on the server. This is compiled from source on Debian 
Squeeze, 32bit, in case that might matter.

Here's the complete conf file, with non-public info elided:

set mail-format {
from: address@hidden
message: TRU-DB1-PRIMARY $SERVICE $EVENT
}
# Don't check during db backup (latency too high)
check program servercomm-db1 with path 
/usr/local/bin/check_server_wrapper with timeout 25 seconds
not every "40-50 0 * * *"
alert address@hidden on with reminder on 4 cycle
alert address@hidden with reminder on 4 cycle
alert address@hidden on with reminder on 4 cycle
alert address@hidden on with reminder on 4 cycle
if status != 0 then alert


--
Paul Theodoropoulos
www.anastrophe.com