Re: [exim] Retry rules don't seem to work in 4.88

Pàgina inicial
Delete this message
Reply to this message
Autor: Dan Liles
Data:  
A: exim-users
Assumpte: Re: [exim] Retry rules don't seem to work in 4.88
Hello Phil,

Thanks for the clarification - it appears centos uses -q1h so a new runner
every hour.

I am also calling exim -q with no other parameters as part of my process of
inserting items in the queue as well as trying exim -odb -M <msgid>.

I hope that answers your questions.

So what do you guys think I should do to diagnose this?

What I think is happening is that the connection is being dropped by google
with out a response so it's not getting logged as an attempt which is
forcing another retry and basically it get's in a loop trying to send the
same message over and over again.

Should I try a packet sniffer to look at all the packets being sent when a
message gets in this state? I've also noticed an event for tcp connect and
smtp connect that I could add some logging to but depending on where the
issue lies it could be hard to try to diagnose relying on exims logging.

Anyway - any pointers, suggestions or comments you could provide on how to
track this issue down would be wonderful.

I'll let you know what I find.

Dan


-----Original Message-----
From: Phil Pennock [mailto:pdp@exim.org]
Sent: Thursday, February 2, 2017 1:27 PM
To: Dan Liles <danliles72@???>
Cc: exim-users@???
Subject: Re: [exim] Retry rules don't seem to work in 4.88

On 2017-02-02 at 10:54 -0600, Dan Liles wrote:
> Thank you so much for your replies. I've got answers below for all your

questions and comments.

> I've looked at Mike's recommendations for command line options but

[...]


My question was about how it is started and explicitly mentioned the -q*
flags.

For instance, if I send my running daemon a SIGHUP so that it restarts, then
it logs:

  2017-02-02 19:22:15 [56860] exim 4.88 daemon started: pid=56860, -q2m,
                              listening for SMTP on [... snip ...]


See that "-q2m"? That's the thing telling Exim that it needs to
periodically start queue-runners, to go through the queue and try
deliveries. If you don't start the daemon with that option, then you need
to start queue-runners yourself instead; cron is the usual method for that.

Eg, if you start Exim from inetd, then you'll start queue-runners from cron.
If you run Exim as a stand-alone daemon, then you'll usually tell it to
start queue-runners itself.

-Phil