From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hannes Reinecke Subject: Re: RHEL6.2: path failures during good path I/O Date: Mon, 18 Jun 2012 13:16:28 +0200 Message-ID: <4FDF0E0C.1090403@suse.de> References: <4FD87335.3040300@linux.vnet.ibm.com> <20120613131613.GA18293@redhat.com> <4FDA3840.9030307@linux.vnet.ibm.com> <20120614211928.GA30587@redhat.com> <4FDEFCFB.70202@linux.vnet.ibm.com> <4FDF0AAB.7000001@linux.vnet.ibm.com> Reply-To: device-mapper development Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: <4FDF0AAB.7000001@linux.vnet.ibm.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dm-devel-bounces@redhat.com Errors-To: dm-devel-bounces@redhat.com To: Christian May Cc: dm-devel@redhat.com List-Id: dm-devel.ids On 06/18/2012 01:02 PM, Christian May wrote: > So, I've started on my RHEL6.1 system filesystem I/O against 5 > multipath devices. Each multipath device contains 4 partition - two > of them were mounted and used for fs-I/O. The exerciser was running > for appr. 2 hours without a single path failure message. Then I'v > started block I/O against the other 5 multipath devices. > Pretty soon the first path failure was reported: > = Could be a request-queue starvation. As you're using directio any I/O requests from the checker will be queued onto the request queue. If the system is fully loaded it might take some time for the checker I/O to be actually submitted; occasionally this might be longer than the I/O timeout of the checker. I would switch to 'tur' checker and retest. (Well, _actually_ I would switch to _SLES_ and retest, but I guess that's beside the point :-) Cheers, Hannes -- = Dr. Hannes Reinecke zSeries & Storage hare@suse.de +49 911 74053 688 SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 N=FCrnberg GF: J. Hawn, J. Guild, F. Imend=F6rffer, HRB 16746 (AG N=FCrnberg)