From owner-freebsd-stable@FreeBSD.ORG Sun Apr 6 09:29:26 2003 Return-Path: Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id D236937B401 for ; Sun, 6 Apr 2003 09:29:26 -0700 (PDT) Received: from netmint.com (netmint.com [207.106.21.130]) by mx1.FreeBSD.org (Postfix) with ESMTP id E9AF043FBD for ; Sun, 6 Apr 2003 09:29:25 -0700 (PDT) (envelope-from support@netmint.com) Received: from netmint.com (localhost.netmint.com [127.0.0.1]) by netmint.com (8.12.6/8.12.5) with ESMTP id h36GTOp7013141 for ; Sun, 6 Apr 2003 12:29:24 -0400 (EDT) (envelope-from support@netmint.com) Received: from localhost (support@localhost) by netmint.com (8.12.6/8.12.3/Submit) with ESMTP id h36GTO6h013138 for ; Sun, 6 Apr 2003 12:29:24 -0400 (EDT) Date: Sun, 6 Apr 2003 12:29:24 -0400 (EDT) From: Support To: stable@freebsd.org Message-ID: <20030406120112.I11095-100000@netmint.com> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Subject: load testing and tuning a 4GB RAM server X-BeenThere: freebsd-stable@freebsd.org X-Mailman-Version: 2.1.1 Precedence: list List-Id: Production branch of FreeBSD source code List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Sun, 06 Apr 2003 16:29:27 -0000 Hello everyone, First of all, great job on the 4.8-R. We have been a long standing user of FreeBSD and are very happy with everything. Now my question. I am trying to stress test a new Dell PowerEdge server and find the limits of its hardware and my tuning. Here are the server stats: * 2x Xeon 2.8 with SMP compiled, hyperthreading NOT compiled in kernel * 4 GB of RAM, 8 GB of swap on Raid 1 (15K RPM drives) for speed * User Data is on Raid 5 in a detached PowerVault Contents of /boot/loader.conf: kern.ipc.maxsockets=32768 kern.ipc.nmbclusters=32768 kern.ipc.nmbufs=131072 kern.ipc.nsfbufs=6656 kern.ipc.shm_use_phys=1 kern.maxfiles=32768 kern.maxproc=8192 kern.maxswzone=33554432 kern.nbuf=16384 kern.ncallout=32768 kern.vm.kmem.size=268435456 kern.vm.pmap.shpgperproc=2048 net.inet.tcp.tcbhashsize=16384 Contents of /etc/sysctl.conf: kern.ip.somaxconn=4096 vm.v_free_min=131072 vm.v_free_target=262144 vm.f_free_reserved=32768 vm.v_free_severe=65536 kern.ps_showallprocs=0 Some info from kernel config: maxusers 0 options KVA_PAGES=512 options NSWAPDEV=1 (the rest is pretty much my hardware, firewall and other irrelevant stuff) This has been copied by hand as the machine is not networked yet, so please excuse any possible typos. The system does a make -j8 world in about 15 minutes. Make -j500 buildworld takes about an hour or less, and the server load skyrockets to 200 or so. Now here's the kicker. I started 7,500 (yes 7.5 K :) lynx processes to eat up available memory and started doing the -j500 buildworld. The machine went into swap about a gig but everything completed fine. It didn't even seem sluggish in controls. I feel I am not pushing it enough. Unfortunately, setting -j over 500 doesn't add any load. If I randomly kill and start 1000s of processes at once, with the total running hovering around 8000 at all times, I can push the "load" on this machine to ridiculous numbers like 700. It can do that while doing a stressful buildworld. I guess the good thing is this machine safely goes into swap when exhausting its 4 gigs and that tells me everything is OKay. sysctl -a |grep kvm shows about 2 GB of memory allocated to kernel and somewhere between 0.7 GB to 0.9 GB free depending on whether the system just booted or is running 8000 processes. Any ideas? Comments? Is my tuning decent, or would you change anything else? This server is to be a very loaded web/database server. I will be happy to hear your thoughts. Please cc: me on all messages because I am not presently subscribed to the list. Thanks again guys, Andrew NetMint, LLC