Hi - I've been trying to use supervisord to start uwsgi while still letting uwsgi run its own awesome process manager.
I've seen other people mention doing this, and also seen some conflicting advice. for me it starts fine, but if I stop them: supervisorctl stop uwsgi not only do the processes not die but they get very pissed off and bang all the CPUs to 100% and sit there screaming. its a staging server, so this is mostly just amusing I think the uwsgi process manager won't allow it to quit. supervisor thinks they are dead because the pid went dead main uwsgi pid is given either QUIT TERM or KILL all of them fail anybody else know about this ? should I just give up and put it in upstart or init ? here is the supervisor conf: [program:uwsgi] command = /usr/sbin/uwsgi --ini /home/crucial/html/nsb/var/conf/ns-uwsgi.conf directory = /home/crucial/html/nsb/ user = crucial priority = 1 stopwaitsecs = 60 stopsignal = QUIT stdout_logfile = /home/crucial/html/nsb/var/log/uwsgi.log stderr_logfile = /home/crucial/html/nsb/var/log/uwsgi.log note: stopwaitsecs/stopsignal means it sends QUIT, waits 60 seconds and then does KILL initially I had the default which is TERM and then KILL if it doesn't respond same problem either way ns-uwsgi.conf: [uwsgi] socket = /tmp/blahblahblah.com.sock chmod-socket = true pythonpath = /home/crucial/html/nsb/bin wsgi-file = bin/django.wsgi single-interpreter = false processes = 8 #workers = 5 #threads = 16 harakiri = 30 master = 1 post-buffering = 8192 post-buffering-bufsize = 65536 -grazie
_______________________________________________ uWSGI mailing list [email protected] http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
