Re: Question

[rory]

OK, I'll go ahead and fix this now that I know it's not my code doing this...

This may take a few days - I'll have to think about this one. (It's been
driving me nuts for a day and a half)
But, the good news is that it looks like I'm monitrc order independant
just be refactoring existing functions in control.c.
I also pulled out the ability for control.c to send stop/start via http.
It stopped making sense to me to do it that way. One way, we called a
function that then called spawn, the other way we sent data to the http
port, which then calls spawn. By removing this in control.c, evrything
runs through check_process() now, including cervlet and validate, and most
of the logic can be put there, instead of replicated in 2 or 3 places.
> Yeah - i though about it too (it is monit's default behavior). It will
> be needed to implement it so, that it will keep state of p->do_validate
>  after reload. Current behavior will forget its value after
> configuration  reload. It is bad for cluster 'manual' mode too -
> monit's configuration  reload is dangerous in such case. We maybe
> shouldn't release 3.1 before  this problem will be solved.
>
> Martin
>
> address@hidden wrote:
>
>>Yes, but if I run "monit stop foo" it should stay stopped, and the
>>do_validate should not be set back to true when I re-load the
>>configuration.
>>
>>So, the second question is:
>>
>>I didn't change my config file, so what triggers a config reload?
>>
>>
>>
>>
>>>Rory Toma wrote:
>>>
>>>
>>>
>>>>Let me clarify. I thought that the code that set the defaults in p.y
>>>>was only called on monit startup. What conditions could cause me to
>>>>visit this code again?
>>>>
>>>>On Wed, 2002-11-20 at 19:13, Rory Toma wrote:
>>>>
>>>>
>>>>
>>>>
>>>>>So, I've about got the order dependency nailed. However, I've broken
>>>>>validate in doing so. I've tracked my problem down to p.y. For some
>>>>>reason, the way I'm doing it is invoking the code in p.y that sets
>>>>>the default value of do_validate to TRUE, clobbering the FALSE value
>>>>>I've put in.
>>>>>
>>>>>Hopefully I can figure a way around this tomorrow. If not, I'll post
>>>>>the patch and see if one of you can.
>>>>>
>>>>>
>>>>>
>>>>>
>>>Configuration reload
>>>
>>>Martin
>>>
>>>
>>>
>>>
>>>_______________________________________________
>>>monit-dev mailing list
>>>address@hidden
>>>http://mail.nongnu.org/mailman/listinfo/monit-dev
>>>
>>>
>>
>>
>>
>>
>>
>>_______________________________________________
>>monit-dev mailing list
>>address@hidden
>>http://mail.nongnu.org/mailman/listinfo/monit-dev
>>
>>
>
>
>
>
>
> _______________________________________________
> monit-dev mailing list
> address@hidden
> http://mail.nongnu.org/mailman/listinfo/monit-dev