Hello,
the clean way is to restart the process via monit:
/usr/bin/monit restart myprocess
or at least unmonitor the service before the restart and enable monitoring
again when the start finished:
/usr/bin/monit unmonitor myprocess
<restart the process
/usr/bin/monit monitor myprocess
If the process is restarted externally, monit is not notified that the restart
- if monit will detect that the process is not running in such situation, it
will try to restart the process as if it failed. The restart counter is
internal to monit restart action.
Regards,
Martin
On Nov 17, 2011, at 1:55 PM, YUM System Administration wrote:
> Hi all.
>
> We are using monit to monitor some processes running as daemons but as
> part or as extension (gem) of the ruby-on-rails framework.
>
> Sometimes these daemons die and monit restarts them reliably. But
> sometimes we run into timeout events, i.e. when the process gets
> restarted for a certain amount of times within a given time period.
>
> My question is: Does monit track, if a restart was done by monit itself
> or does it not matter, if the restart was done by someone else / has
> been invoked by a user from shell or something?
> Does monit just count how often a restart has been done within a given
> period of time? If the threshold is reached, the "TIMOUT" Event is
> triggered and monit's surveillance of the daemon is stopped?
>
> The problem is: Every time our developers deploy the rails application,
> the daemons get restarted too. Sometime this happens very frequently (
> every two minutes).
>
> We are running monit 5.1.1 from debian squeeze and 4.10.1 from debian lenny.
>
> Regards
> Hans Ulrich
>
>
> --
> To unsubscribe:
> https://lists.nongnu.org/mailman/listinfo/monit-general
--
To unsubscribe:
https://lists.nongnu.org/mailman/listinfo/monit-general
