Re: [exim] Timeouts and retry rules

Top Page
Delete this message
Reply to this message
Author: Dean Brooks
Date:  
To: exim-users
Subject: Re: [exim] Timeouts and retry rules
On Thu, Sep 02, 2004 at 04:29:26PM +0100, Philip Hazel wrote:

> > Problem is, each individual message gets retried EVERY time a queue
> > run goes over each message. It's as though no retry rule ever
> > gets entered. I've appended the search for a particular message
> > from our mainlog today.
> >
> > 2004-09-02 06:25:26 1Bxu0L-000737-K6 == admin@??? R=mwdirect
> > T=remote_smtp defer (145): Connection timed out: SMTP timeout while
> > connected to 64.253.97.107 [64.253.97.107] after end of data (3341
> > bytes written)
>
> 1. What is in your retry database for this host, if anything? (Search
> with exinext.)


Here's the output for the above message:

T:64.253.97.107:64.253.97.107:1Bxu0L-000737-K6 145 65 Connection timed out
19-Aug-2004 17:09:25 02-Sep-2004 11:18:29 02-Sep-2004 15:18:29

What's odd is that this same message had been attempted 4 minutes
earlier at 11:14:29, which would have presumably set a 4-hour retry
time, but it didnt - it reattempted at 11:18:29.

> 2. Try a delivery with debugging enabled; see if it says anything about
> retry data.


Ok, I did a "-q 1Bxu0L-000737-K6 1Bxu0L-000737-K6" to preserve retry
data on the run.

checking status of 64.253.97.107
locking /var/spool/exim/mailworks/db/retry.lockfile
locked /var/spool/exim/mailworks/db/retry.lockfile
opened hints database /var/spool/exim/mailworks/db/retry: flags=0
dbfn_read: key=T:64.253.97.107:64.253.97.107
dbfn_read: key=T:64.253.97.107:64.253.97.107:1Bxu0L-000737-K6
no host retry record
host+message retry time not reached: checking ultimate address timeout
address match: subject=*@64.253.97.107 pattern=*
64.253.97.107 in "*"? yes (matched "*")
*@64.253.97.107 in "*"? yes (matched "*")
on queue longer than maximum retry for address - allowing delivery
64.253.97.107 [64.253.97.107] status = usable

So it appears the maximum retry issue is in play. However, from the docs,
it would seem that it would bounce the message on the timeout.

But, I get this in the debug log after it times out:

   == admin@??? R=mwdirect T=remote_smtp defer (145):
      Connection timed out: SMTP timeout while connected to 64.253.97.107
      [64.253.97.107] after end of data (3341 bytes written)
  locking /var/spool/exim/mailworks/db/retry.lockfile
  locked /var/spool/exim/mailworks/db/retry.lockfile
  opened hints database /var/spool/exim/mailworks/db/retry: flags=2
  address match: subject=*@64.253.97.107 pattern=*
  64.253.97.107 in "*"? yes (matched "*")
  *@64.253.97.107 in "*"? yes (matched "*")
  retry for T:64.253.97.107:64.253.97.107:1Bxu0L-000737-K6 (eosystems.net) = *
  dbfn_read: key=T:64.253.97.107:64.253.97.107:1Bxu0L-000737-K6
  Writing retry data for T:64.253.97.107:64.253.97.107:1Bxu0L-000737-K6
    first failed=1092949765 last try=1094139700 next try=1094154100 expired=0
    error 145 65: Connection timed out
  dbfn_write: key=T:64.253.97.107:64.253.97.107:1Bxu0L-000737-K6
  end of retry processing


--
Dean Brooks
dean@???