Date: Fri, 1 Nov 1996 14:44:17 +0100 (MET) From: Mikael Karpberg <karpen@ocean.campus.luth.se> To: jgreco@brasil.moneng.mei.com (Joe Greco) Cc: greg@uswest.net, freebsd-hackers@freefall.freebsd.org, ponds!rivers@dg-rtp.dg.com Subject: Re: Another data point in the daily panics... Message-ID: <199611011344.OAA08663@ocean.campus.luth.se> In-Reply-To: <199610311645.KAA27895@brasil.moneng.mei.com> from Joe Greco at "Oct 31, 96 10:45:48 am"
next in thread | previous in thread | raw e-mail | index | archive | help
According to Joe Greco: > > > Recall, this is a 386DX-33, 8MEG, IDE controller, sio ports, running > > > my news and mail and SL/IP connections. > > I run several stable news servers too (soon to be more)... Mr. Rivers > seems to be running a "stress test from hell" configuration :-( I read some mail someone showed me, not long ago... I think it was from the NetBSD lists... Some guy stating (in some argument with a Linux hacker) that "Your machine is probably idle while you read this", upon which the Linux hacker said "No, it's not. I fire up 40 creashmes in init". I'd say that linux hacker is boooored ;-) Then again, I know a friends 2.1.5 machine rebooted after running 30 seconds of "crashme". Firing up _40_ at boot time, just for kicks, must mean he can get a lot more out of Linux's vm/fs system then we can get out of FreeBSD's, when it comes to stablilty. If you want to stress the machine, and shake out bugs, "crashme" seems quite a nice stresser for the system, doing a lot of mean stuff, but nothing illegal (which should not be possible anyway, or the system is not very safe, or?) as far as I know. I know we got HUGE amounts of "sig 10 recieved" when running crashme. So, if crashme is not reading longs out of alignment, or so, then there is a problem in the system somewhere. Anyone tried to fire up 40 crashmes and wait? Should produce nice output for debugging a stressed system, no? /Mikael
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199611011344.OAA08663>