Date: Sat, 20 Oct 2012 08:49:17 -0700 From: Dennis Glatting <freebsd@penx.com> To: Andriy Gapon <avg@freebsd.org> Cc: freebsd-fs@freebsd.org Subject: Re: ZFS hang status update Message-ID: <1350748157.88577.15.camel@btw.pki2.com> In-Reply-To: <50825598.3070505@FreeBSD.org> References: <1350698905.86715.33.camel@btw.pki2.com> <1350711509.86715.59.camel@btw.pki2.com> <50825598.3070505@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
--=-T3qbQpcvQzfHgpDvLbi3 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit On Sat, 2012-10-20 at 10:41 +0300, Andriy Gapon wrote: > on 20/10/2012 08:38 Dennis Glatting said the following: > > This is da0 (the cache --SSD) on which camcontrol hanged. It is on the > > same controller. > > Hmm, hanging camcontrol is a bad sign. It would be interesting to get procstat > -k information just for the hanging camcontrol process. > Also, is it possible to eliminate this disk from the configuration? > Attached (I hope but also found here: http://www.pki2.com/zfs_stats_efficiency-week.png) is a munin graph of "ZFS ARC Efficiency - by week" for the server I have been talking about. You will notice before each crash the prefetch efficiency is going down. This graph is for daily graph. http://www.pki2.com/zfs_stats_efficiency-day.png --=-T3qbQpcvQzfHgpDvLbi3--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1350748157.88577.15.camel>