Re: Question

[rory]

So, probably the only thing that it could be is that I'm somehow stomping
on monitrc's timestamp/checksum data, causing a reload.
In any case, I'll keep the code that makes a copy of processlist before
blowing it away, and sets p->do_validate correctly.
I can probably get everything functional tomorrow, but it'll probably be
next week before I get something nice and cleaned up to check in.
> On Thu, 21 Nov 2002 address@hidden wrote:
>
>> > There is the only one cause to reload monit's configuration without
>> > monit restart - changed timestamp of monitrc. If you'll change
>> > configuration and immediately load it, you can send monit signal to
>> > wakeup. It then looks on timestamp and parses monitrc again. There
>> > isn't  any other way (to force configuration reload) now. In the
>> > past i made  patch for monit implementing SIGHUP, that caused forced
>> > configuration  reload, but it was rejected (it isn't realy needed).
>> >
>> > Martin
>> >
>>
>> I wonder what else would do it. I've managed to code myself into a
>> place where this is happening, and I don't know why.
>
> If you grep over the whole code with "do_[a-z]*init" you see that
> "do_init" and "do_reinit" are only called in exactly one place.
> do_init happes in main and only at startup.  And do_reinit is started
> in do_default and it can only be called if "monitrc" has changed.
>
> As long as there is no problem with is_rcfile_changed(), there can't be
> any unexpected config reloads.
>
> Christian
>
> --
> Christian Hopp                                email:
> address@hidden Institut für Elektrische Informationstechnik
>            fon: +49-5323-72-2113 TU Clausthal, Leibnizstr. 28, 38678
> Clausthal-Zellerf.   fax: +49-5323-72-3197
>                             pgpkey:
>                             https://www.iei.tu-clausthal.de/pgp-keys/
>
>
>
>
> _______________________________________________
> monit-dev mailing list
> address@hidden
> http://mail.nongnu.org/mailman/listinfo/monit-dev