On Tue, Oct 29, 2024 at 02:07:56PM +0900, Nico Schottelius via Postfix-users
wrote:
> Viktor Dukhovni via Postfix-users <postfix-users@postfix.org> writes:
>
> > On Sat, Oct 26, 2024 at 12:06:12AM +0900, Nico Schottelius via
> > Postfix-users wrote:
> >
> >> The maps/hashes that make a lot of sense on VMs/servers for avoiding
> >> reloading postfix, do not make much sense in the k8s/container context.
> >
> > Restarts are much more disruptive that reloads, because the entire
> > active queue moves back to the incoming queue, client connections are
> > closed abruptly, ... It is not clear why k8s makes that a non-issue.
>
> First of all, I fully agree with you in regards to reload vs. restart
> being disruptive, if we are considering a single system.
A Postfix restart is disruptive on a running system that has client
connections and an active queue. The client connections are terminated
and the queue is substantially reshuffled. If the system is not already
running, then it isn't a "restart".
This remains true in clusters, unless what you're describing is stateful
process migration, not a stop/start (restart).
--
Viktor.
_______________________________________________
Postfix-users mailing list -- postfix-users@postfix.org
To unsubscribe send an email to postfix-users-le...@postfix.org
