Date: Sat, 26 Jun 1999 11:46:36 +0400 From: Alex Povolotsky <tarkhil@asteroid.svib.ru> To: aphor@ripco.com Cc: stable@freebsd.org Subject: Re: Panic: ffs_clusteralloc: map mismatch Message-ID: <199906260746.LAA21649@shuttle.svib.ru> In-Reply-To: Your message "Fri, 25 Jun 1999 18:51:04 CDT." <377415E8.D24E8DCE@ripco.NOSPAM.com>
next in thread | previous in thread | raw e-mail | index | archive | help
<377415E8.D24E8DCE@ripco.NOSPAM.com>Jeremy McMillan writes:
>The first thing I noticed is that Alex is using softupdates. I assume at
>this point that we are dealing with a softupdates problem, and the error
>"looks" like ffs may have had bad assumptions about the state of things. In
>any case, I'm watching this list closely for softupdates issues, so if this
>is a softupdates thing please don't take the discussion off-list.
Well, if the panic is caused by softupdates, it's REALLY BAD THING.
Softupdates gave me huge (x2 on sam tasks) performance boost, but I've got a
drawback also. I guess it happens only on IDE systems, I haven't met this
disaster on SCSI.
Any process actively using disk (starting exmh or netscape are the most common
of them) causes all other processes literelly to halt. Even response to enter
pressed in shell delays by about 1-2 seconds.
I have posted my DMESG already, so I see no point in repeating.
Until I've started to burn CDs, it didnt' hurt me REALLY, but yesterday it
costed me 3 disks and about 30 minutes.
Any ideas, anyone?
I'm using 3.2-19960619,
/usr/src/contrib/sys/softupdates/ffs_softdep.c:
$Id: ffs_softdep.c,v 1.20.2.2 1999/05/12 02:13:03 dg Exp $
/usr/src/contrib/sys/softupdates/softdep.h:
$Id: softdep.h,v 1.4.2.1 1999/05/12 02:13:04 dg Exp $
Alex.
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199906260746.LAA21649>