Re: [Assp-test] Antwort: Re: Antwort: Re: Antwort: Re: Antwort: Re: Antwort: Re:ASSP V2 -ready tobereleased on SF ?

[GrayHat <gr...@gm...>]

>> won't be a somewhat
>> better approach having the web GUI running in a separate thread ?

> Yes Yes Yes !

:)

>> well... the GUI thread may also handle this... although I'm not sure
it
>> would be possible since the thread would then need to access
>> data/infos which sit inside the main process... hmmm....

> This is the realy problem. I want to have such a WebThread and my
> early plans (2007/2008) have seen it. The real problem is the shutdown
> and restart - which is impossible to do outside the main thread and
> having control on it.

mumble.... handling such a thing in "C" would mean having a global
mutex or semaphore which the threads should poll and which would
be used to signal the shutdown/reboot; not sure if/how this may be
possible in perl though; also, and since we're at it, when dealing with
shared data, my usual approach (C again) is to create the needed
number of "marshaling" functions; then the various thread call those
functions which in turn handle the access to the common data and
manage all the cuncurrency stuff avoiding race conditions; the main
advantage is that the threads don't need to know about the inner
details of data access (acquire a mutex or the like) and, at the same
time, thanks to common data pool the memory fingerprint is reduced;
again, not sure if or how it may be done in perl but maybe it's worth
a thought

> I think it would be much more easy, to switch the
'SMTP-connection-listening'
> to an other/extra thread - and keep Web and Stats in the MainThread.
> We need some news for the new releases - so let's keep this in mind.

agreed; moving the listener to its own thread may probably be easier
and will allow to keep the ASSP email processing running even if the
main thread is handling ... something else

>> Jul-6-09 18:43:16 [Main_Thread] Info: Worker_4 : last sigoff in main,
> c:\assp\assp.pl, 21827, main::sigoffTry, 1, , ,  at Jul-6-09 18:29:12
> 1246897752.3445 - 21827

> This looks like ClamAV is not responding - holding the socket in
> C-code and Perl is unable to get it out there ! Try to restart ClamAV
> - if no success - restart ASSP.

hmm... ClamAV was working and even responding (tested it with clamdscan)
tried to restart it, restarted fine, still, ASSP is logging those
messages every
time it gets an SMTP connection, sounds like some connections get stuck
and are only cleaned when ASSP gets a new connection (not sure but it's
what I'm guessing by looking at the "live" log); I'm going to stop /
restart
ASSP now ... stopping ...

<============================================================>
Jul-7-09 09:09:59 [Main_Thread] Service stopping
Jul-7-09 09:09:59 [Main_Thread] Waiting for high Workers to be finished
Jul-7-09 09:10:02 [Worker_10001] Worker_10001 finished
Jul-7-09 09:10:03 [Worker_10000] Worker_10000 finished
Jul-7-09 09:10:03 [Main_Thread] High workers finished work
Jul-7-09 09:10:03 [Main_Thread] Saving whitelist
Jul-7-09 09:10:03 [Main_Thread] Info: 7893 records of Whitelist saved
Jul-7-09 09:10:03 [Main_Thread] Saving redlist
Jul-7-09 09:10:03 [Main_Thread] Info: 47 records of Redlist saved
Jul-7-09 09:10:03 [Main_Thread] Saving delaying records
Jul-7-09 09:10:04 [Main_Thread] Info: 17598 records of Delay saved
Jul-7-09 09:10:04 [Main_Thread] Info: 63 records of DelayWhite saved
Jul-7-09 09:10:04 [Main_Thread] Saving ldaplist
Jul-7-09 09:10:04 [Main_Thread] Info: 1713 records of LDAPlist saved
Jul-7-09 09:10:04 [Main_Thread] Saving penalty records
Jul-7-09 09:10:04 [Main_Thread] Info: 6009 records of PBBlack saved
Jul-7-09 09:10:05 [Main_Thread] Info: 1690 records of PBWhite saved
Jul-7-09 09:10:05 [Main_Thread] Info: 1 records of PBTrap saved
Jul-7-09 09:10:05 [Main_Thread] Saving cache records
Jul-7-09 09:10:05 [Main_Thread] Info: 32 records of RBLCache saved
Jul-7-09 09:10:05 [Main_Thread] Info: 73 records of URIBLCache saved
Jul-7-09 09:10:05 [Main_Thread] Info: 639 records of SPFCache saved
Jul-7-09 09:10:05 [Main_Thread] Info: 133 records of PTRCache saved
Jul-7-09 09:10:05 [Main_Thread] Info: 98 records of MXACache saved
Jul-7-09 09:10:05 [Main_Thread] Info: 616 records of SBCache saved
Jul-7-09 09:10:05 [Main_Thread] Info: 699 records of RWLCache saved
Jul-7-09 09:10:05 [Main_Thread] Info: 2 records of DKIMCache saved
Jul-7-09 09:10:05 [Main_Thread] Info: 0 records of BATVTag saved
Jul-7-09 09:10:21 [Main_Thread] Info: 442947 records of BackDNS saved
Jul-7-09 09:10:21 [Main_Thread] Info: saving Stats in file asspstats.sav
Jul-7-09 09:10:21 [Worker_2] Worker_2 finished
Jul-7-09 09:10:21 [Worker_3] Worker_3 finished
Jul-7-09 09:10:21 [Worker_1] Worker_1 finished
Jul-7-09 09:10:21 [Main_Thread] Waiting for all SMTP-Workers to be
finished
Jul-7-09 09:10:21 [Worker_5] Worker_5 finished
Jul-7-09 09:11:11 [Main_Thread] Info: saving Stats in file asspstats.sav
Jul-7-09 09:11:11 [Main_Thread] ASSP finished work
Jul-7-09 09:11:12 [Main_Thread] Service stopped

<============================================================>

ok, restarting now... restarted, and those "sigoff" messages
disappeared; I've
to say that upon first startup yesterday I had no such messages they
started
appearing after some time; so I wonder if there's something "piling up"
and
causing ASSP to log them; again, ClamD seems to be fine so I'd look at
something else... although don't know what :)

Ok... will let it run and report later; now gotta run away for some
hours

HTH