From owner-freebsd-hackers Wed Aug 26 11:36:09 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id LAA20274 for freebsd-hackers-outgoing; Wed, 26 Aug 1998 11:36:09 -0700 (PDT) (envelope-from owner-freebsd-hackers@FreeBSD.ORG) Received: from hoth.ffwd.bc.ca ([209.153.243.242]) by hub.freebsd.org (8.8.8/8.8.8) with SMTP id LAA20267 for ; Wed, 26 Aug 1998 11:36:07 -0700 (PDT) (envelope-from skye@ffwd.bc.ca) Received: from skye by hoth.ffwd.bc.ca with local (Exim 1.82 #3) id 0zBkOq-00006y-00; Wed, 26 Aug 1998 11:34:28 -0700 Message-ID: <19980826113428.29710@ffwd.bc.ca> Date: Wed, 26 Aug 1998 11:34:28 -0700 From: Skye Poier To: freebsd-hackers@FreeBSD.ORG Subject: Re: VM crash? questions pt.2 References: <19980820172445.54455@ffwd.bc.ca> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.85 In-Reply-To: <19980820172445.54455@ffwd.bc.ca>; from Skye Poier on Thu, Aug 20, 1998 at 05:24:45PM -0700 X-URL: http://www.ffwd.bc.ca/ Sender: owner-freebsd-hackers@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Update..... Thanks to suggestions from Nick Himba I tried the following: # dd if=/dev/rwd0s1b of=/dev/null Did that several times, no errors. Got 1,153,740 bytes/sec Then I booted into single-user mode (swap partition not swapon'd) and tried: # dd if=/dev/urandom of=/dev/rwd0s1b Again, no errors. It was incredibly slow though! 122,064 bytes/sec I thought urandom may be a bit slow so I tried # dd if=/dev/zero of=/dev/rwd0s1b And got 197,955 bytes/sec. Pretty cruddy for a Quantum Fireball in Mode 4 on a fairly new EIDE controller, no? But anyway, on with the program... Then I tried newfs'ing the swap partition. No errors, ran a couple fsck's. Then I mounted the partition on /mnt and did: # dd if=/dev/urandom of=/mnt/bigfile And bingo, the computer locked up. I rebooted (off floppy, more about that below) back into single-user and remounted the swap part after fscking it, the size of /mnt/bigfile was around 31 MB (the swap partition is 65MB). Couple things I noticed: when my AMD 486/120 went to CPU heaven I put in an Intel DX4-66 but I don't think I got the jumpers quite right because the BIOS reports it running at 80MHz. A long shot, but I'll fix that and see if it helps any. Second thing, when I crashed the machine by creating /mnt/bigfile it seems to have smacked the BSD bootloader :P I have a "dangerously dedicated" partitioning scheme, how can I put the bootloader back in place? Somewhere in /stand/sysinstall? Luckily I have a 2.2.6 boot disk, I booted the kernel from its bootloader with boot: 0:wd(0,a)kernel Thanks, Skye Previously I said: > I am running 2.2.6-RELEASE on a 486 with 16mb of memory and a 65mb swap > partition. > > Recently I've been working with some rather large files and the machine > has had a tendancy to lock up solid. No error on the console, in dmesg, > nothing. It seems to happen when VM (via swapinfo) reaches 30% usage or > about 19MB of the 65MB avaiable.... other than this the server has been > totally bulletproof. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message