From owner-freebsd-hackers Thu Jun 15 21:03:17 1995 Return-Path: hackers-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id VAA18136 for hackers-outgoing; Thu, 15 Jun 1995 21:03:17 -0700 Received: from cancer.ibms.sinica.edu.tw ([140.109.40.250]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id VAA16765 for ; Thu, 15 Jun 1995 21:00:38 -0700 Received: (from taob@localhost) by cancer.ibms.sinica.edu.tw (8.6.11/8.6.9) id MAA04704; Fri, 16 Jun 1995 12:00:23 +0800 Date: Fri, 16 Jun 1995 12:00:21 +0800 (CST) From: Brian Tao To: Mark Hittinger cc: hackers@freebsd.org Subject: re: Too many open files in system In-Reply-To: <199506151344.JAA21312@ns1.win.net> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: hackers-owner@freebsd.org Precedence: bulk On Thu, 15 Jun 1995, Mark Hittinger wrote: > > I use: > > maxusers 256 > child_max=128 > open_max=128 > nmbclusters=768 > > swap space on disk = 128mb Trying that config now (except with only 64 megs swap on a 16-meg system). > I don't see the problem as "often". Could some other parameter be keyed > off of the maxusers that influences this? In addition I use a different > web server (cern), so maybe there is a file descriptor leak in the one > you are using? It very well could be the problem. I haven't been following NCSA httpd's progress too closely, but I just downloaded 1.4.1 last night (been using 1.4). Even with the above options, the machine/VM pager still chokes up and dies. However, if I substitute the Apache 0.65 server (standard forking model) for NCSA httpd, the problem goes away. I'll try NCSA in standard forking mode next to see if that is the problem. Unfortunately, I no longer have any pre-2.0.5 machines around for an additional comparison. :-/ -- Brian ("Though this be madness, yet there is method in't") Tao taob@gate.sinica.edu.tw <-- work ........ play --> taob@io.org