Re: [exim] Exim 4.94 - daemon_notifier_socket bind: Address …

Top Page
Delete this message
Reply to this message
Author: David Carter
Date:  
To: Jeremy Harris
CC: exim-users
Subject: Re: [exim] Exim 4.94 - daemon_notifier_socket bind: Address already in use
On 2020-07-03 12:30, Jeremy Harris via Exim-users wrote:

>> today I gave Exim 4.94 a try and ran into the same problem. I must
>> admit that my configuration requires to run two Exim instances in
>> parallel. I'm executing the following two commands to start the
>> daemons:
>> [...]
>>
> I guess this is really a bug, but unless you (or someone) gives a
> good need for multiple daemons it's low priority.


We ran into this as well. In our case the two daemons are:

/opt/exim/bin/exim -bd -q5m
/opt/exim/bin/exim -q1m -Rr [@.]cam.ac.uk$ -oP
/spool/exim/exim-q-hermes.pid

The second daemon forces frequent retries for local email (1m rather
than 5m) to stop a backlog forming if a local mail system is down for a
short period of time.

daemon_notifier_socket() seems to be new in 4.94: it wasn't there in
4.92.3

Ultimately the paniclog entries seem to be harmless and I can ignore
them (or use -oX), but this is a separate use case for multiple daemons.