> What is the recommended way to use a plugin + spooler. I don't know if
> this is an issue or a feature. If I use the example C plugin, add a
> spooler function, and on the uwsgi commandline remap identifier1 from 250
> to 0, then some spooler is called for 0 and and mine is called for 250.
>
> I get around this by settting the identifier1 to 0 in my "real" app as I
> will select this particular instance of uwsgi by an explicit route prefix.
> Thus, I am fine with the current implementation, but I am noting what I
> found in the event that it is helpful for someone else to know.
>
> --Jeff
> _______________________________________________
> uWSGI mailing list
> [email protected]
> http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
>
The idea is that every plugin is scanned for a spooler hook.
After having called the hook the following code is called:
// hook returns SPOOL_IGNORE
if (ret == 0) continue;
callable_found = 1;
// hook returns SPOOL_OK
if (ret == -2) {
uwsgi_log("[spooler] done with task %s after %d seconds\n", task,
time(NULL)-now);
destroy_spool(dir, task);
}
// hook returns SPOOL_RETRY
// ...re-spool it
break;
This allows to have (for example) a ruby and a perl app both having their
spooler manager.
--
Roberto De Ioris
http://unbit.it
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
