httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Jeff Trawick <>
Subject Re: [PATCH] convert worker MPM to leader/followers design
Date Thu, 11 Apr 2002 17:08:14 GMT
"Bill Stoddard" <> writes:

> > -----
> > > Jeff Trawick wrote:
> > >
> > > >Aaron Bannert <> writes:
> > > >
> > > >>I appreciate that you are trying to moderate my usage of the -1 (veto),
> > > >>but feel it is my duty to inform the list as soon as possible that
> > > >>wouldn't be happy with this big of a change.
> > > >>
> > > >
> > > >Maybe a veto wasn't appropriate, but it seemed clear to me that a
> > > >number of people would rather have the new code side by side with the
> > > >old for a while, and Brian didn't seem disturbed about that notion at
> > > >all.  Hopefully nobody will actually go remove worker as soon as
> > > >Brian's code is committed :)
> > > >
> > >
> > > I disagree.  Unless someone wants to volunteer to put a workaround
> > > in the current worker code to fix the queue-full case, I can't
> > > rationalize including it in another beta or GA release.  We need
> > > to either fix it or remove it.
> > >
> > > --Brian
> >
> > I am in grudging agreement with Brian. I don't have a lot of time to spend on this,
> > I'll take a look at making worker behave more like mpm_winnt, which should fix the
> > problem.
> >
> > Bill
> Before anyone goes weird on me for that last comment, I am specifically referring to
> mpm_winnt manages the queue between the accept and worker threads.  Not planning on
> changing anything else about how worker operates other than how the accept/worker queue
> managed.

I'll be interested in exactly what you want to change.  I'd be shocked
if this problem can't be solved without more than tweaking the design.
Graceful termination (after user intervention or maxrequests or some
syscall failure) is built around the current queue handling, so it is
valuable to avoid changing the current code too much, particularly if
people have ideas about a better overall design which could be ready
before too long.

We have this recurring problem with the MPMs where somebody wants to
switch out the design for some MPM function because the new design
supposedly doesn't have some problem that the current design has.
Just about every time new problems are introduced and not resolved and
whoever did the design switch-out is disinterested or busy.

Something else I "go weird on" is the idea that worker isn't good
enough to go in another beta or GA release.  It is in pretty decent
shape right now.  prefork is the default MPM anyway.  Look at
perchild, for crying out loud.  And how many people read and work on
PRs? There's stuff that doesn't work right with any MPM.  Should that
function be removed before another beta or GA release unless the PR is
resolved?  (I don't think so.)

Jeff Trawick |
Born in Roswell... married an alien...

View raw message