[Mailman-Users] (no subject)
adamsca at gmail.com
Tue May 14 18:28:40 CEST 2013
Mark, I was reading some previous posts about Mailman and qfiles. Late
yesterday, I found a couple of *.bak files in the out/ queue, so I moved
all the *.bak and *.pck files to another directory. I restarted Mailman.
Mail was not still not being delivered in a timely way.
Then, this morning, I read another post about dns lookups in Postfix main.cf.
A poster said that he modifed his smtpd_recipient_restrictions and put
permit_mynetworks at the top. So, I followed this and restarted Postfix.
I immediately saw mail flowing and the out/q queue emptied and mail began
to come in. I'm not sure exactly why it had that affect.
So, I have a final question. Should I move the .pck and .bak (renamed to
.pck) files back to the out/ queue. If so, should I stop mailman before
BTW - The posts that I read above involved you. so thanks for the
On Tue, May 14, 2013 at 8:26 AM, Christopher Adams <adamsca at gmail.com>wrote:
> Thank you for your reply.
> QRUNNER_SLEEP_TIME = seconds(1) in Defaults.py
> Mesages are being archived almost immediately. There is nothing in the in/
> There currently are about 954 files in the out/ queue and seemingly always
> growing. Is that significant?
> A message sent yesterday at 12:30 was received by Mailman within seconds.
> It delivered at 3:30. The smtp.log for that message shows:
> May 13 15:32:36 2013 (26531)
> <0B03FE68E0E760478D413F045B39A028380323EA at OSLEXCHANGE.my.mail.server.local>
> smtp to test7 for 2 recips, completed in 0.020 seconds
> On Mon, May 13, 2013 at 9:03 PM, Mark Sapiro <mark at msapiro.net> wrote:
>> On 05/13/2013 08:39 PM, Christopher Adams wrote:
>> > Postfix sends the list message it to the remote server, which shows
>> > it was accepted by the Mailman server.
>> We are not communicating here. The above makes no sense to me. I think
>> the sequence is:
>> - List mail arrives at the local Postfix.
>> - Postfix delivers to the local Mailman.
>> - Mailman delivers - here I don't know if to the local Postfix for
>> relay to the remote server or directly to the remote server. The
>> mm_cfg.py setting for SMTPHOST determines this. The Default is localhost
>> meaning relay via the local Postfix.
>> > The Mailman server has logs that
>> > show that the message was directed to the command 'post' to list
>> > However, nothing happens after that for 2-7 hours. Once it actually
>> > to the list, there are smpt.log entries that show it going to 'x
>> OK, so here you are saying that the delay is definitely in Mailman. I.e.
>> Postfix delivers to the mailman mail wrapper which presumably
>> immediately queues the message in Mailman's in/ queue, and it's a long
>> time later that Mailman logs the delivery.
>> What is the processing time in the smtp log entry. Is it a few seconds
>> or less or hours?
>> I'm guessing that during this time IncomingRunner or OutgoingRunner or
>> both are asleep. Do you have a setting in mm_cfg.py for
>> QRUNNER_SLEEP_TIME? The default in Defaults.py is (or should be)
>> QRUNNER_SLEEP_TIME = seconds(1)
>> If you have set this to a few hours, it would explain the behavior you
>> are seeing. If not, when does the post get archived? Is that delayed too?
>> Also, messages in the in/ queue are processed by IncomingRunner and then
>> queued in the archive/ queue for ArchRunner and the out/ queue for
>> OutgoingRunner. You could look and see if messages move from the in/
>> queue quickly or not.
>> Note that if you have set QRUNNER_SLEEP_TIME to say 4 hours, messages
>> can sit in the in/ queue for an average of 2 and up to 4 hours before
>> moving to the out/ queue where they may wait up to an additional 4 hours
>> before being sent.
>> Mark Sapiro <mark at msapiro.net> The highway is for gamblers,
>> San Francisco Bay Area, California better use your sense - B. Dylan
> Christopher Adams
> adamsca at gmail.com
adamsca at gmail.com
More information about the Mailman-Users