[Rspamd-Users] redis not shutting down

[Randy Bush]

>> debian 13, current postfix and rspamd.  if i need to reboot to get a new
>> kernel or whatever, shutdown hangs looping trying to shut redis down.
>> the net of a million lies says to use the hard reboot hammer.  this
>> seems a bit brutal and possibly damaging to data.  is there a better
>> path?
> 
> I'm *guessing* that systemd is causing the problem.  Not unusual in my
> sometimes painful experiences with systemd.  I once had systemd killing
> fsck every 90 seconds at boot.  The fsck in question took a few minutes
> (it was before I moved from ext3 to ext4) so that didn't go too well.
> 
> Disabling the service as far as systemd is concerned may tell you if
> systemd is in fact the problem.  To use redis you would then have to
> start it some other way, e.g. manually after stopping it or a reboot.
> 
> My approach *if* this is the problem would be to remove redis from the
> systemd configuration entirely so that it can't get in the way, and use
> some other way to start and stop redis.  Others might tell you that you
> can tweak the systemd configuration to fix it e.g. increasing timeouts,
> but I'm on the hanging bench - at least as far as systemd is concerned.

i hate systemd and ipv6 as much as the next dinosaur, but i prefer to be
a bit pragmatic

i sudpect that the issue is that the systemd config for redis, at least
in trixie, has `TimeoutStopSec=0`, and `systemctl edit resdis.service`
to prepend, for example, `TimeoutStopSec=60`, should have the desired
effect

randy