Mailing-List: contact cvs-help@httpd.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@httpd.apache.org
Date: 26 Apr 2002 22:45:47 -0000
Message-ID: <20020426224547.24232.qmail@icarus.apache.org>
From: brianp@apache.org
To: httpd-2.0-cvs@apache.org
Subject: cvs commit: httpd-2.0 STATUS

brianp      02/04/26 15:45:47

  Modified:    .        STATUS
  Log:
  Added a vote on MPMs
  
  Revision  Changes    Path
  1.606     +10 -2     httpd-2.0/STATUS
  
  Index: STATUS
  ===================================================================
  RCS file: /home/cvs/httpd-2.0/STATUS,v
  retrieving revision 1.605
  retrieving revision 1.606
  diff -u -r1.605 -r1.606
  --- STATUS	26 Apr 2002 17:43:30 -0000	1.605
  +++ STATUS	26 Apr 2002 22:45:47 -0000	1.606
  @@ -1,5 +1,5 @@
   APACHE 2.0 STATUS:                                              -*-text-*-
  -Last modified at [$Date: 2002/04/26 17:43:30 $]
  +Last modified at [$Date: 2002/04/26 22:45:47 $]
   
   Release:
   
  @@ -83,7 +83,15 @@
   
       * Make the worker MPM the default MPM for threaded Unix boxes.
         +1:   Justin, Ian, Cliff, BillS
  -      -0:   Lars
  +      -0:   Lars,
  +            BrianP (according to the system call profiles, worker's main
  +            bottleneck--and the reason why it underperforms compared to
  +            prefork--is in mutex ops, and there's no easy way to fix
  +            that without turning it into something that isn't really worker
  +            any more...I think we need to either switch to leader/follower
  +            (and do enough testing and debugging to make the L/F code
  +            production-quality) or come up with another design that has
  +            less mutex contention than worker)
         -1:   Aaron (premature decision, needs more discussion and bug fixes)
   
   RELEASE NON-SHOWSTOPPERS BUT WOULD BE REAL NICE TO WRAP THESE UP: