Re: [PATCH] Total-#-of-starts counter

[Martin Pala]

Hi Michel,

this is nice and simple patch (and you did good job with ntp test :) 
however i think it is too specific for service restart.

Monit creates events based on testing rules and service condition. This 
involves specific action as defined by the rule. The 'restart' action is 
just one type - you can see them all in monitor.h:

#define ACTION_ALERT       1
#define ACTION_RESTART     2
#define ACTION_STOP        3
#define ACTION_EXEC        4
#define ACTION_UNMONITOR   5
#define ACTION_START       6

In addition to the action type it is important to know the related 
timeframe. The total start's counter makes no diferrence between 10 
restarts during last 10 minutes and between 10 restarts during last 
year. So, the patch introduces basic event history - the total start 
counter (special case of monit action).

I think it is better to separate such functionality from monit. The new 
m/monit version which we work on has this functionality integrated 
already. It allows to collect all types of events from all monitored 
nodes, sort them by various criteria, etc. For illustration is in the 
attachment screenshot of related m/monit part (just basic screenshot - 
it has much more functionality). It can also make graphs of runtime 
service properties, such as space usage, load average, etc.

So, my vote is -1 for this patch.


Thanks,

Cheers,
Martin


Michel Marti wrote:
> Hello,
>
> Attached is a patch against monit-CVS that adds a (re)start counter to a 
> service. It is increased every time that a service is (re)started and 
> reset if the service unmonitored. The counter is displayed on the 
> html/text/xml status page.
>
> Can you apply this patch?
>
>
> Cheers,
>
> Michel