Hi there, Since I switched to TCP sockets on my uwsgi installation I noticed some strange things in the netstat output. Let's say we have uwsgi running, with --pidfile /var/run/app.pid -s 127.0.0.1:25555. If /var/run/app.pid contains a PID like 26712 I would expect to find a netstat entry like this:
tcp 0 0 127.0.0.1:25555 0.0.0.0:* LISTEN 26712/uwsgi However after some days running uwsgi with probably some harakiri going on, I found this: tcp 0 0 127.0.0.1:25555 0.0.0.0:* LISTEN 609/sh Note that the PID is wrong, and the command as well. I looked at this process and found: 1003 609 0.0 0.0 4024 584 ? S Sep17 0:00 /bin/sh -c apachesyslog Actually this is one of the subprocesses that are spawned from my uwsgi workers... On another host I found the PID of another worker in the netstat output, not the master! Apparently I cannot use the output of netstat reliably. Any idea what's going wrong? -- Jean-Baptiste Quenot _______________________________________________ uWSGI mailing list [email protected] http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
