Justin Erenkrantz wrote:
>
> On Thu, Jul 19, 2001 at 12:03:16PM -0400, Greg Ames wrote:
> > What about the threaded mpm? I can't tell you how it's doing in
> > production, unfortunately. Yeah, I know a number of folks have changes
> > they'd like to see in it, but as far as I know it's in better shape than
> > it was in our 2.0.16 beta.
> >
> > tag?
>
> +0 (not that my vote is binding in this tree).
>
> I'm also not sure of the impact from inherit in the APR code has been
> tested yet. Wouldn't a pipe (i.e. POD) need to have the inherit flag
> explicitly set?
The PoD is working correctly, since graceful restart and
perform_idle_server_maintenance are both working.
>
> I think it'd be nice if we shipped a decent default threaded config
> (i.e. don't have it spawn 3 child processes and kill off 2 immediately
> because the config is lame) and fix the POD reading in threaded.
> Neither are showstoppers though (the POD one may cause end-user
> problems though).
>
> I probably wouldn't pay attention to threaded MPM until we have the SMS
> code active by default. Once we've removed the per-process allocation
> mutex, we can start to focus on getting other things right in there.
Again, let's compare the current cvs tree (including the threaded mpm)
to our one and only 2.0 beta. I believe the current code wins hands
down, or am I missing something?
Yeah, we have bugs to fix and improvements to make. But there are folks
out there who need threaded and want a beta/binaries. As I recall,
there were enough bugs with threaded in 2.0.16 so that I wouldn't
recommend it to anybody who can only deal with binaries.
Greg
