Le Fri, 9 Aug 2013 21:20:01 -0400, Alex Riina <ari...@flipkey.com> a écrit : > I'd like to start migrating processes over to use supervisor, but I'm > hoping for a little more control over the processes that I have found > so far. Is there a canonical way to handling "soft restarts" or tear > downs in supervisor processes? For example, if I have a process like: > > setup_logic() > while True: > run_logic() > > Is there a recommended way to schedule a restart of the process > without interrupting the current loop? Is there a way to add a > teardown hook to the process?
AFAICT, a common way to handle soft restarts (aka "reloads") of a daemon is to send it a SIGHUP (and for the daemon to intercept the signal, of course). > For the second question, I'm currently registering my teardown via > signal.signal(signal.SIGTERM, teardown), but I feel like both must be > such common questions that better solutions must exist. I'm afraid not :-) Regards Antoine. _______________________________________________ Supervisor-users mailing list Supervisor-users@lists.supervisord.org https://lists.supervisord.org/mailman/listinfo/supervisor-users
