Milter-greylist

Emmanuel Dreyfus wrote:

>
>> In order to get the database saved when the process gets killed, I've
>> added a few lines to catch SIGTERM (which the start/stop-script sends)
>> and call exit (which dumps the database via onexit()). You can take
>> a look at it here:
>>
>> 	ftp://ftp.nada.kth.se/Sysgr/milter.signal.patch
>
> I'm not sure we can do that: libmilter documentation explicitely says
> it takes care of all signals. Catching the signal may break the 
> library API
> in the future.


I understand your concern. However without this, the database will
not get dumped on Solaris 9 when the milter is stopped. I guess there
should be some callback in the libmilter (maybe the xxfi_abourt()?) to
handle this.

  I'll keep the patch around maybe #ifdef:ed so it works as it should
in Solaris 9.


Is 1.5.8 considered as stable as 1.5.3? Could I run this on a production
system (and thus test the peer-patch)?


/klas

Klas Heggemann <klas@...> wrote:

> I understand your concern. However without this, the database will
> not get dumped on Solaris 9 when the milter is stopped. I guess there
> should be some callback in the libmilter (maybe the xxfi_abourt()?) to
> handle this.

I've checked the termination on signal, it dumps a core (without dumping
the database) on NetBSD-2.0 because a locked mutex is destroyed. I'll
fix that and we'll see if the problem remains.
 
> Is 1.5.8 considered as stable as 1.5.3? Could I run this on a production
> system (and thus test the peer-patch)?

I can't swear anything, but it should be alright. If it's not it's
important someone discover the problem, anyway :)

-- 
Emmanuel Dreyfus
Il y a 10 sortes de personnes dans le monde: ceux qui comprennent 
le binaire et ceux qui ne le comprennent pas.
manu@...