From: Aurelien Barrere (no email)
Date: Fri Sep 02 2005 - 10:11:02 EDT
Greg Hackney wrote:
> Aurelien,
>
> A couple of thoughts... are most of the messages mailing-list bounces
> that
> get pushed through your perl bounce engine? If so, I see that they
> are limited
> to one-at-a-time via "local_destination_concurrency_limit = 1".
> Could that be a
> bottleneck ?
> relay=local, delay=357272, status=sent (delivered to
> command: /opt/bounceengine/bounces.pl
>
> Also, you don't mention what hardware or OS you are running. On some
> really older systems, I've
> seen slowdowns in Postfix delivery after directory sizes and numbers
> grew huge for the queue area (which
> was fixed by recreated the queue directories and moving the files back
> into it).
> hash_queue_depth = 2
> --
> Greg Hackney
> Cypress, Texas
>
>
> Aurelien Barrere wrote:
> relay=local, delay=357272, status=sent (delivered to command:
> /opt/bounceengine/bounces.pl
>
>> John Peach wrote:
>>
>>> On Thu, 01 Sep 2005 18:06:18 +0200
>>> Aurelien Barrere <> wrote:
>>>
>>>
>>>
>>>> Hi all,
>>>>
>>>> Since one month I noticed that my queue never decreases
>>>> I begin to panic because it s become huge like a blackhole.
>>>> Since one month I monitored regulary the queue and it 's started at
>>>> 278.000 and now one month after I ve got more than 6.000.000 in the
>>>> queue !
>>>>
>>>>
>>>
>>>
>>>
>>> ...and what do your logs tell you?
>>>
>>>
>>>
>> Here is a snapshot (/var/log/maillog):
>>
>> Sep 1 18:08:34 bounce postfix/qmgr[29484]: 87FCE76C3CB2: from=<>,
>> size=30083, nrcpt=1 (queue active)
>> Sep 1 18:08:34 bounce postfix/smtpd[19036]: 506E77915EA7:
>> client=ftl1smtp01.citco.com[208.219.172.67]
>> Sep 1 18:08:34 bounce postfix/smtpd[17433]: 5742B7920B86:
>> client=vg1.soc2.com[62.23.133.253]
>> Sep 1 18:08:34 bounce postfix/cleanup[20284]: 5742B7920B86:
>> message-id=<>
>> Sep 1 18:08:34 bounce postfix/smtpd[18579]: 5ED037970617:
>> client=omr-m11.mx.aol.com[64.12.138.23]
>> Sep 1 18:08:34 bounce postfix/smtpd[17433]: 74C9E7A52E58:
>> client=vg1.soc2.com[62.23.133.253]
>> Sep 1 18:08:34 bounce postfix/cleanup[19606]: 506E77915EA7:
>> message-id=<B7F244FD679C8D4E998E795B95E59DB10170F468 at syd1msex01>
>> Sep 1 18:08:34 bounce postfix/cleanup[18644]: 74C9E7A52E58:
>> message-id=<>
>> Sep 1 18:08:34 bounce postfix/smtpd[16642]: connect from
>> omr-m05.mx.aol.com[64.12.138.17]
>> Sep 1 18:08:34 bounce postfix/local[20447]: 86115798A4C9:
>> to=<>, orig_to=<>, relay=local,
>> delay=357272, status=sent (delivered to command:
>> /opt/bounceengine/bounces.pl)
>> Sep 1 18:08:34 bounce postfix/qmgr[29484]: 86115798A4C9: removed
>> Sep 1 18:08:34 bounce postfix/qmgr[29484]: 870D67AA4202: from=<>,
>> size=5810, nrcpt=1 (queue active)
>> Sep 1 18:08:34 bounce postfix/smtp[19165]: 87F6F7AA3F16:
>> to=<>,
>> relay=gateway02.mercer.com[192.12.251.37], delay=175635, status=sent
>> (250 Message received and queued)
>> Sep 1 18:08:34 bounce postfix/qmgr[29484]: 87F6F7AA3F16: removed
>> Sep 1 18:08:34 bounce postfix/qmgr[29484]: 87FA6792C5E7: from=<>,
>> size=1840, nrcpt=1 (queue active)
>> Sep 1 18:08:34 bounce postfix/smtpd[19036]: disconnect from
>> ftl1smtp01.citco.com[208.219.172.67]
>> Sep 1 18:08:34 bounce postfix/smtpd[20252]: A3CB5798A4C9:
>> client=www1.hagenhosting.com[63.97.115.201]
>> Sep 1 18:08:34 bounce postfix/cleanup[19600]: 5ED037970617:
>> message-id=<>
>> Sep 1 18:08:34 bounce postfix/smtpd[17433]: AA2607A52FC2:
>> client=vg1.soc2.com[62.23.133.253]
>> Sep 1 18:08:34 bounce postfix/cleanup[19606]: AA2607A52FC2:
>> message-id=<>
>> Sep 1 18:08:34 bounce postfix/smtpd[16642]: BB98D7A5305A:
>> client=omr-m05.mx.aol.com[64.12.138.17]
>> Sep 1 18:08:34 bounce postfix/smtpd[17433]: C7A827A5322B:
>> client=vg1.soc2.com[62.23.133.253]
>> Sep 1 18:08:34 bounce postfix/cleanup[19605]: C7A827A5322B:
>> message-id=<>
>> Sep 1 18:08:34 bounce postfix/smtpd[18579]: disconnect from
>> omr-m11.mx.aol.com[64.12.138.23]
>> Sep 1 18:08:34 bounce postfix/local[20375]: 866CA7848BC8:
>> to=<>, orig_to=<>, relay=local,
>> delay=357269, status=sent (delivered to command:
>> /opt/bounceengine/bounces.pl)
>> Sep 1 18:08:34 bounce postfix/qmgr[29484]: 866CA7848BC8: removed
>>
>> ---
>> According to me, It seems that relay=local works well and after that
>> message is removed after being proceed by perl
>>
>> Hope that this snapshot can be useful...
>> AB
>
>
>
>
2 things:
1/
I modified in master.cf according to tuning chapter
(http://www.postfix.org/TUNING_README.html)
smtp inet n - n - - smtpd
to
smtp inet n - n - 10 smtpd
and I obtained this message from netsaint:
Socket timeout after 10 seconds
Thus I decided to restore the default value
2/
It becomes BETTER since i change the value you suggested me:
from "local_destination_concurrency_limit = 1" to
"local_destination_concurrency_limit = 30"
You can see below that incoming queue begins to decrease :-)
[root at bounce abarrere]# cat -n count_vik_queue.2005-09-02-12h00m.txt
1 maildrop 1
2 incoming 6417664
3 active 20260
4 hold 1
5 deferred 447
[root at bounce abarrere]# cat -n count_vik_queue.2005-09-02-13h41m.txt
1 maildrop 1
2 incoming 6367955
3 active 20262
4 hold 1
5 deferred 447
[root at bounce abarrere]# cat -n count_vik_queue.2005-09-02-15h46m.txt
1 maildrop 1
2 incoming 6347696
3 active 20257
4 hold 1
5 deferred 482
So Thanks for this suggestion
What I notice is I got a warning message in the log saying:
Sep 1 14:32:02 bounce postfix/qmgr[29484]: warning: lots of deferred
mail, that is bad for performance
Ok but what can i do?
Regards
AB
|
|
|