> I thought the idea behind the "2" version of "smart-attach-daemon" was for
> daemon's that didn't create their own PID file.  If the process creates
> it's own PID then we should just use "smart-attach-daemon" and point to
> the
> PID created by that process, no?

"smart" daemons are meant to be run "detached" from uWSGI, that is why we
only need to track the shell (if the main process dies, the shell is
destroyed automatically). The '2' variant, allows uWSGI to track the shell
saving its pidfile, so it can monitor it.

Basically you are giving the pidfile another meaning (while it is only
used internally by uWSGI to track the shell)

-- 
Roberto De Ioris
http://unbit.it
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi

Reply via email to