On Wed, Apr 10, 2002 at 02:03:23PM -0700, Roy T. Fielding wrote: > What Brian describes is a fatal flaw in the current Worker MPM code. > In order to veto his change, you will be expected to review the code > first and then propose a better solution. Until then, choose a better > way than -1 to make reservations. > > If Brian needs to create a new MPM to fix this kind of problem, then > I will remove the worker MPM from the code base as soon as the new one > is committed. I don't believe in leaving code in httpd that is not > robust enough for normal use.
The particular problem that Brian is describing (hanging when one process' worker-queue has filled) was a case that I specifically tested for and did not encounter back when I was working in that code. I have not yet had time to review the current state of the worker MPM, but I believe that it is something that can be corrected in the current model. My veto was intended to imply that I do not want the model of the worker MPM changed, but that I am totally in favor of this new model and would support a new MPM to implement it. 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 I wouldn't be happy with this big of a change. -aaron
