From mboxrd@z Thu Jan 1 00:00:00 1970 From: Mike Christie Subject: Re: RE: Dell MD3000i issues with dm mp Date: Wed, 15 Apr 2009 12:33:29 -0500 Message-ID: <49E61A69.6090807@cs.wisc.edu> References: Reply-To: device-mapper development Mime-Version: 1.0 Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: quoted-printable Return-path: In-Reply-To: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Sender: dm-devel-bounces@redhat.com Errors-To: dm-devel-bounces@redhat.com To: device-mapper development List-Id: dm-devel.ids Moger, Babu wrote: > I don=92t see any apparent reason for these messages(=93rdac Dettached=94= ) by looking at the log. There are some issci session failure messages. = I don=92t understand those messages very well. You may have to ask some = iscsi experts to look at it. >=20 The http://pastebin.com/d97ae1ad link does not work anymore. If you=20 could put the log back up, I can look at the iscsi messages. > ________________________________ > From: dm-devel-bounces@redhat.com [mailto:dm-devel-bounces@redhat.com] = On Behalf Of Kurt Bechstein > Sent: Tuesday, April 14, 2009 12:44 PM > To: device-mapper development > Subject: [dm-devel] RE: Dell MD3000i issues with dm mp >=20 > No devices were detached during this sequence at least not physically a= nyways. >=20 > From: dm-devel-bounces@redhat.com [mailto:dm-devel-bounces@redhat.com] = On Behalf Of Moger, Babu > Sent: Tuesday, April 14, 2009 1:38 PM > To: device-mapper development > Subject: [dm-devel] RE: Dell MD3000i issues with dm mp >=20 > I am seeing few =93rdac Dettached=94 messages. There are only messages = I am bit concerned about (unless you disconnected some devices purposeful= ly during the test). I am not concerned about other messages. >=20 > Thanks > Babu Moger > ________________________________ > From: dm-devel-bounces@redhat.com [mailto:dm-devel-bounces@redhat.com] = On Behalf Of Kurt Bechstein > Sent: Tuesday, April 14, 2009 12:10 PM > To: device-mapper development > Subject: [dm-devel] RE: Dell MD3000i issues with dm mp >=20 > Here is a snippet of my /var/log/messages from this morning=92s boot se= quence after iscsi started and some of the output I=92m seeing. Seems li= ke quite a bit to me. >=20 > http://pastebin.com/d97ae1ad >=20 > From: dm-devel-bounces@redhat.com [mailto:dm-devel-bounces@redhat.com] = On Behalf Of Moger, Babu > Sent: Tuesday, April 14, 2009 12:12 PM > To: device-mapper development > Subject: [dm-devel] RE: Dell MD3000i issues with dm mp >=20 > I would think it should be pretty sparse. >=20 > Thanks > Babu Moger > ________________________________ > From: dm-devel-bounces@redhat.com [mailto:dm-devel-bounces@redhat.com] = On Behalf Of Kurt Bechstein > Sent: Tuesday, April 14, 2009 10:57 AM > To: device-mapper development > Subject: [dm-devel] RE: Dell MD3000i issues with dm mp >=20 > I actually tried doing that by doing a mkinitrd /boot/initrdfile =96pre= load=3Dscsi_dh_rdac kernelversion and booted using the initrd and it stil= l seemed to crank out quite a few errors. With doing it this way, should= I still expect to see a good deal of I/O errors or should they pretty sp= arse? >=20 > From: dm-devel-bounces@redhat.com [mailto:dm-devel-bounces@redhat.com] = On Behalf Of Moger, Babu > Sent: Tuesday, April 14, 2009 11:38 AM > To: device-mapper development > Subject: [dm-devel] RE: Dell MD3000i issues with dm mp >=20 > Yes, this could happen during the boot where the scsi_dh_rdac handler i= s not yet loaded. You might want to try adding scsi_dh_rdac in the initrd= . This will help to cut short all your error messages. I don=92t see any= problem with your multipath.conf. >=20 > Thanks > Babu Moger > ________________________________ > From: dm-devel-bounces@redhat.com [mailto:dm-devel-bounces@redhat.com] = On Behalf Of Kurt Bechstein > Sent: Tuesday, April 14, 2009 9:54 AM > To: dm-devel@redhat.com > Subject: [dm-devel] Dell MD3000i issues with dm mp >=20 > Greetings everyone, I have a RHEL 5.3 host attached to two Dell MD3000i= storage arrays via iscsi. I=92m having some interesting issues upon boo= tup in this setup. Between the two arrays I=92m accessing 24 luns so aft= er my iscsi service kicks in I get tons and tons of I/O error messages go= ing across the screen. This can go on for 15-20 minutes or longer. It w= ould appear that some I/O is being sent to the passive paths and they are= returning errors. It would seem the scsi layer is doing its thing to sc= an for partitions and probably causing the I/O that is being seen. Is th= ere any way to suppress/avoid this scenario? Also, I thought I would pos= t my multipath.conf file just to have some other eyes have a look at it a= nd make sure I=92m not missing something obvious so that is below. Thank= s in advance! >=20 >=20 > devices { > device { > vendor "DELL" > product "MD3000i" > features "0" > getuid_callout "/sbin/scsi_id -g -u -s /block/%= n" > prio_callout "/sbin/mpath_prio_rdac /dev/%n" > hardware_handler "1 rdac" > path_grouping_policy group_by_prio > failback immediate > path_checker rdac > prio "rdac" >=20 > } > } > multipaths { > mulitpath { > device { > vendor DELL > product MD3000i > } > } > } >=20 >=20 >=20 > -----------------------------------------------------------------------= - >=20 > -- > dm-devel mailing list > dm-devel@redhat.com > https://www.redhat.com/mailman/listinfo/dm-devel