Hi Roberto, We're still having the locking issue on 1.1 latest snapshot as well, doesn't seem to resolve unless we use flock.
Should uWSGI use flock by default for everything, or is there something more we can do to resolve this problem? Cal On Tue, Feb 28, 2012 at 8:28 AM, Roberto De Ioris <[email protected]> wrote: > > Il giorno 28/feb/2012, alle ore 08:46, Roberto De Ioris ha scritto: > > > > > Il giorno 28/feb/2012, alle ore 08:40, Evgeny Turnaev ha scritto: > > > >> Ok. I will try to make mmaped posix semaphores in lock.c and > >> python_plugin.c soon. > > > > i am rewriting a bit the locking api. > > Some user may want to have instances-aware locking (you can lock from > multiple uWSGi instances) > > using files or ipc. I think i will commit it in an hour. > >> > > Ok, i have committed the new code. > > Adding new lock engine should be easy. > > You have to only work on lock.c > > Users can request "alternative" lock engines with the --lock-engine option > > On FreeBSD flock is automatically selected > > -- > Roberto De Ioris > http://unbit.it > JID: [email protected] > > _______________________________________________ > uWSGI mailing list > [email protected] > http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi >
_______________________________________________ uWSGI mailing list [email protected] http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
