Rép.

I think my problem is solved!

As Gary V. suggested I checked the postfix log. I found
quite a bit of these
messages:

	Warning: premature end-of-input on private/scache socket
while
reading 	input attribute name

I found that this error was due to Postfix Connection Cache
(h
ttp://www.postfix.org/CONNECTION_CACHE_README.html). 

Then it happened again. 

So I did the following tests

	- Telnet localhost 10025 and sent myself an email: it
worked
	- Telnet localhost 10024 and sent myself an email: it
worked

I was then convinced the problem with Postfix and somewhere
before it sends
mail to amavis.

I noticed that mail was building up in my active queues but
more
specifically one particular domain had a crazy amount of
mail in it. I
noticed that the other domains mail was not moving and
starting to increase
slowly. 

I added the following line to main.cf which overrides the
default setting

smtp_connection_cache_on_demand=no

Immediately all the mail from the domain which add a crazy
amount moved to
the deferred queue and all the other mail went through. 

I then tried to connect to the mail server which postfix was
configured to
relay mail from the abusing domain to. I couldn't the server
was timing out
since it was overloaded (I'm guessing). 

It turns out that all these emails were bounces from a mass
mail out they
had done. 

I have to read more about Connection Caching but basically
this is what was
holding everything up.

If anybody could explain exactly what was going on it would
be appreciated &
maybe useful to the list.

Anyways, I think I need to get to know postfix much much
better... (any
recommendations for reading material???)

Thanks for all your help

Simon

-----Original Message-----
From: maia-users-bouncesrenaissoft.com
[mailto:maia-users-bouncesrenaissoft.com] On Behalf
Of Gary V
Sent: December 20, 2006 10:18 AM
To: maia-usersrenaissoft.com
Subject: RE: [Maia-users] Rép. : Postfix Issue

>Thanks for the reply. I read your replies to the
"How many processes & no
>responce on telnet port 25" thread. I then made the
changes you suggested.
>
>Here is what my master.cf file now:
>
>smtp      inet  n       -       n       -       100    
smtpd -o
>  content_filter=smtp-amavis:[127.0.0.1]:10024

100 is the default, so you could say:
smtp      inet  n       -       n       -       -     smtpd
-o
  content_filter=smtp-amavis:[127.0.0.1]:10024

>
>smtp-amavis unix  -  -  n  -  -  smtp
>     -o smtp_data_done_timeout=1200
>     -o smtp_send_xforward_command=yes
>     -o disable_dns_lookups=yes
>smtp      unix  -       -       n       -       -      
smtp
>
>localhost:10025 inet	n	-	n	-	-	smtpd -o
>content_filter=
>     -o smtpd_restriction_classes=
>     -o smtpd_delay_reject=no
>     -o
smtpd_client_restrictions=permit_mynetworks,reject
>     -o smtpd_helo_restrictions=
>     -o smtpd_sender_restrictions=
>     -o
smtpd_recipient_restrictions=permit_mynetworks,reject
>     -o smtpd_data_restrictions=reject_unauth_pipelining
>     -o smtpd_end_of_data_restrictions=
>     -o mynetworks=127.0.0.0/8
>     -o smtpd_error_sleep_time=0
>     -o smtpd_soft_error_limit=1001
>     -o smtpd_hard_error_limit=1000
>     -o smtpd_client_connection_count_limit=0
>     -o smtpd_client_connection_rate_limit=0
>     -o smtpd_milters=
>     -o local_header_rewrite_clients=
>     -o local_recipient_maps=
>     -o relay_recipient_maps=
>     -o
>receive_override_options=no_header_body_checks,no_unknow
n_recipient_checks
>
>I still get the same problem. Basically, postfix seems
to stop processing
>mail. If I look at the top command while it is stalled
nothing seems to be
>going on.

How much swap memory is used and what does a typical amavisd
process look 
like in terms of memory usage?

>When I run a postfix reload command I see the 10 amavisd
processes
>come up right away. I don't think amavis is the problem
because it answers
>to a telnet request (telnet localhost 10024). I'm really
not sure what to
>do.
>
>I see that you have suggested:
>
>smtp-amavis unix  -       -       n       -       2     
 smtp
>
>I have:
>
>smtp-amavis unix  -  -  n  -  -  smtp
>
>But I have the following in my main.cfg file
>
>smtp-amavis_destination_concurrency_limit = 10
>

smtp-amavis unix  -       -       n       -       2      
smtp
Is only an example. This would be for someone who has
$max_servers at 2.

Since you have left the maxproc at default:
smtp-amavis unix  -  -  n  -  -  smtp

you are correct in setting this in main.cf:
smtp-amavis_destination_concurrency_limit = 10

It has the same effect as setting:
smtp-amavis unix  -  -  n  -  10  smtp

You should have $max_servers set to 10 so they match.

>I would greatly appreciate help on this.
>
>Thanks again for the replies
>
>Simon

I don't have a SuSE 10 system to play with but I would start
by seeing what 
Postfix errors there are.
egrep '(warning|error|fatal|panic):' /var/log/mail

I dont know where amavis logs are stored, but I would also
look for amavis 
errors in the log. The logs are typicall the first place to
look when there 
are problems.
http://www.renaissoft.com/pipermail/maia
-users/2006-December/008478.html

Gary V

____________________________________________________________
_____
Your Hotmail address already works to sign into Windows Live
Messenger! Get 
it now 
http://clk.atdmt.com/MSN/go/msnnkwme
0020000001msn/direct/01/?href=http://get
.live.com/messenger/overview

_______________________________________________
Maia-users mailing list
Maia-usersrenaissoft.com
http://www.renaissoft.com/mailman/listinfo/maia-users
_______________________________________________
Maia-users mailing list
Maia-usersrenaissoft.com
http://www.renaissoft.com/mailman/listinfo/maia-users

Rép.

Thread: Rép.