Date: Sat, 07 Aug 1999 21:00:32 +0200 From: Tor.Egge@fast.no To: joe@nall.com Cc: lightningweb@hotmail.com, freebsd-stable@FreeBSD.ORG Subject: Re: continued crashes with 3.1-Stable Message-ID: <199908071900.VAA51775@midten.fast.no> In-Reply-To: Your message of "Sat, 07 Aug 1999 07:52:02 -0500" References: <37AC2BF2.C4C60F1E@nall.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> > (da0:dpt0:0:0:0): Invalidating pack [...] > > FreeBSD 3.1-STABLE #1 [....] > > - Six 9.1GB Quantum VikingII SCSI3 U2W drives [...] > Don't discount the hardware problem response. We use big (200GB+) [...] I had similar problems when using 3.1-STABLE and 18 GB Quantum Atlas III disks with old firmware (<QUANTUM QM318000TD-SCA N1K0>). A disk dropped the scsi bus, the following selection operation timed out, and the CAM layer immediately invalidated the disk pack. Afterwards, the buffer cache was filled with dirty buffers destined for the bad disk. When the number of free buffers dropped to 0, the machine got stuck in an infinite kernel loop at splbio() protection. See PR 11697 for more details. By having the patch in PR 11697 applied, I was able to avoid the hang, but a reboot was still needed to access the bad disk again. Support for selection retries were added to the CAM code 1999/05/09. This reduces the problem as long as the disks only take short lunch breaks. I've not experienced this problem with newer revisions of the Quantum Atlas III Disk firmware (XN495). Also note that 3.1-STABLE had some problems with low memory hangs which were fixed about 1999/03/18. - Tor Egge 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?199908071900.VAA51775>