Description of problem: Running older version of supervisor which has a broken unix HTTP server socket. Version-Release number of selected component (if applicable): 2.1-8.el6 How reproducible: Install supervisor and try to use the supervisorctl command. Steps to Reproduce: 1. Install supervisor (sudo yum install supervisor) 2. Ensure the /etc/supervisord.conf has a configuration for unix HTTP server: [unix_http_server] file=/var/run/supervisor.sock 3. Start supervisor (sudo service supervisord start) 4. Attempt to invoke the supervisorctl utility (sudo supervisorctl status) Actual results: error: <class 'socket.error'>, [Errno 2] No such file or directory: file: <string> line: 1 Expected results: worker1 RUNNING worker2 STARTING worker3 STOPPED Additional info: The socket file doesn't appear to get created for the unix HTTP server: [vagrant@worker-vagrant ~]$ sudo ls /var/run/supervisor.sock ls: cannot access /var/run/supervisor.sock: No such file or directory There does not appear to be any major functional differences from 2.1 to 3.0 other than many added configuration options (I'm using 2.1 configurations on 3.0). Can we upgrade the EPEL package to the recently released 3.0 version? I was able to rebuild the RPM to the 3.0 version with minimal changes (added explicit /etc/supervisord.conf to init script, require python-setuptools).
The [unix_http_server] section is part of supervisor 3: http://supervisord.org/upgrading.html#upgrading-supervisor-2-to-3 You must configure the http_port option in the [supervisord] section of the config file. The default config file: [supervisord] http_port=/var/tmp/supervisor.sock ; (default is to run a UNIX domain socket server) Closing this ticket... 7 years later :-D