Hello, The bad news is I lost a drive on my production server. The good news is the spare kicked in and everything kept running as it should. I stopped by the datacenter last night, shut down the server, and physically replaced the failed drive. I want to make sure I don''t screw anything up so would somebody be so kind as to lay out what my next steps are? Below is a snippet of what zpool status is showing: tank DEGRADED 0 0 0 mirror-0 DEGRADED 0 0 0 spare-0 DEGRADED 0 0 0 c10t0d0 REMOVED 0 0 0 c10t6d0 ONLINE 0 0 0 c11t0d0 ONLINE snip... spares c10t6d0 INUSE currently in use How do to bring the replaced drive back online and get it into the array? Do I make the new drive the spare or do I bring the new drive online in the mirror and return the original spare to spare status? Any advice and/or actual commands would be greatly appreciated! Thank you, Chris Dunbar
Hi Chris, Which Solaris release is this? Depending on the Solaris release, you have a couple of different options. Here''s one: 1. Physically replace original failed disk and detach the spare. A. If c10t0d0 was the disk that you physically replaced, issue this command: # zpool replace tank c10t0d0 # zpool clear tank Then check zpool status to see if the spare detached automatically. B. If zpool status says that c10t0d0 is ONLINE but the spare is still INUSE, just detach the spare: # zpool detach tank c10t6d0 The next one depends on your Solaris release. Thanks, Cindy On 07/22/11 12:35, Chris Dunbar - Earthside, LLC wrote:> Hello, > > The bad news is I lost a drive on my production server. The good news is the spare kicked in and everything kept running as it should. I stopped by the datacenter last night, shut down the server, and physically replaced the failed drive. I want to make sure I don''t screw anything up so would somebody be so kind as to lay out what my next steps are? Below is a snippet of what zpool status is showing: > > tank DEGRADED 0 0 0 > mirror-0 DEGRADED 0 0 0 > spare-0 DEGRADED 0 0 0 > c10t0d0 REMOVED 0 0 0 > c10t6d0 ONLINE 0 0 0 > c11t0d0 ONLINE > > snip... > > spares > c10t6d0 INUSE currently in use > > > How do to bring the replaced drive back online and get it into the array? Do I make the new drive the spare or do I bring the new drive online in the mirror and return the original spare to spare status? Any advice and/or actual commands would be greatly appreciated! > > Thank you, > Chris Dunbar > _______________________________________________ > zfs-discuss mailing list > zfs-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
Cindy, Thank you for the reply. I am running Solaris Express 11 (5.11 snv_151a i86pc i386 i86pc Solaris). I have physically replaced the drive, but I have not partitioned it yet. I know there is a command to copy the layout from one disk to another and that has worked well for me in the past. I just have to find the command again. Once that is done, do I need to detach the spare before I run the replace command or does running the replace command automatically bump the spare out of service and put it back to being just a spare? Thanks! Chris ----- Original Message ----- From: "Cindy Swearingen" <cindy.swearingen at oracle.com> To: "Chris Dunbar - Earthside, LLC" <cdunbar at earthside.net> Cc: zfs-discuss at opensolaris.org Sent: Friday, July 22, 2011 3:57:00 PM Subject: Re: [zfs-discuss] Replacing failed drive Hi Chris, Which Solaris release is this? Depending on the Solaris release, you have a couple of different options. Here''s one: 1. Physically replace original failed disk and detach the spare. A. If c10t0d0 was the disk that you physically replaced, issue this command: # zpool replace tank c10t0d0 # zpool clear tank Then check zpool status to see if the spare detached automatically. B. If zpool status says that c10t0d0 is ONLINE but the spare is still INUSE, just detach the spare: # zpool detach tank c10t6d0 The next one depends on your Solaris release. Thanks, Cindy On 07/22/11 12:35, Chris Dunbar - Earthside, LLC wrote:> Hello, > > The bad news is I lost a drive on my production server. The good news is the spare kicked in and everything kept running as it should. I stopped by the datacenter last night, shut down the server, and physically replaced the failed drive. I want to make sure I don''t screw anything up so would somebody be so kind as to lay out what my next steps are? Below is a snippet of what zpool status is showing: > > tank DEGRADED 0 0 0 > mirror-0 DEGRADED 0 0 0 > spare-0 DEGRADED 0 0 0 > c10t0d0 REMOVED 0 0 0 > c10t6d0 ONLINE 0 0 0 > c11t0d0 ONLINE > > snip... > > spares > c10t6d0 INUSE currently in use > > > How do to bring the replaced drive back online and get it into the array? Do I make the new drive the spare or do I bring the new drive online in the mirror and return the original spare to spare status? Any advice and/or actual commands would be greatly appreciated! > > Thank you, > Chris Dunbar > _______________________________________________ > zfs-discuss mailing list > zfs-discuss at opensolaris.org > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20110722/e38f5daf/attachment.html>
Chris, You don''t need to partition the drive because your pool contains whole disks. The zpool replace command will do any relabeling that is needed. Do the replace first and then detach the spare, if needed. The spare should detach automatically but sometimes it doesn''t. Thanks, Cindy On 07/22/11 14:12, Chris Dunbar - Earthside, LLC wrote:> Cindy, > > Thank you for the reply. I am running Solaris Express 11 (5.11 snv_151a > i86pc i386 i86pc Solaris). > > I have physically replaced the drive, but I have not partitioned it yet. > I know there is a command to copy the layout from one disk to another > and that has worked well for me in the past. I just have to find the > command again. Once that is done, do I need to detach the spare before I > run the replace command or does running the replace command > automatically bump the spare out of service and put it back to being > just a spare? > > Thanks! > Chris > > ------------------------------------------------------------------------ > *From: *"Cindy Swearingen" <cindy.swearingen at oracle.com> > *To: *"Chris Dunbar - Earthside, LLC" <cdunbar at earthside.net> > *Cc: *zfs-discuss at opensolaris.org > *Sent: *Friday, July 22, 2011 3:57:00 PM > *Subject: *Re: [zfs-discuss] Replacing failed drive > > Hi Chris, > > Which Solaris release is this? > > Depending on the Solaris release, you have a couple of > different options. > > Here''s one: > > 1. Physically replace original failed disk and detach > the spare. > > A. If c10t0d0 was the disk that you physically replaced, > issue this command: > > # zpool replace tank c10t0d0 > # zpool clear tank > > Then check zpool status to see if the spare detached > automatically. > > B. If zpool status says that c10t0d0 is ONLINE but the > spare is still INUSE, just detach the spare: > > # zpool detach tank c10t6d0 > > The next one depends on your Solaris release. > > Thanks, > > Cindy > > > On 07/22/11 12:35, Chris Dunbar - Earthside, LLC wrote: > > Hello, > > > > The bad news is I lost a drive on my production server. The good news > is the spare kicked in and everything kept running as it should. I > stopped by the datacenter last night, shut down the server, and > physically replaced the failed drive. I want to make sure I don''t screw > anything up so would somebody be so kind as to lay out what my next > steps are? Below is a snippet of what zpool status is showing: > > > > tank DEGRADED 0 0 0 > > mirror-0 DEGRADED 0 0 0 > > spare-0 DEGRADED 0 0 0 > > c10t0d0 REMOVED 0 0 0 > > c10t6d0 ONLINE 0 0 0 > > c11t0d0 ONLINE > > > > snip... > > > > spares > > c10t6d0 INUSE currently in use > > > > > > How do to bring the replaced drive back online and get it into the > array? Do I make the new drive the spare or do I bring the new drive > online in the mirror and return the original spare to spare status? Any > advice and/or actual commands would be greatly appreciated! > > > > Thank you, > > Chris Dunbar > > _______________________________________________ > > zfs-discuss mailing list > > zfs-discuss at opensolaris.org > > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss > >
On Fri, Jul 22, 2011 at 1:12 PM, Chris Dunbar - Earthside, LLC <cdunbar at earthside.net> wrote:> I have physically replaced the drive, but I have not partitioned it yet. I > know there is a command to copy the layout from one disk to another and that > has worked well for me in the past. I just have to find the command again. > Once that is done, do I need to detach the spare before I run the replace > command or does running the replace command automatically bump the spare out > of service and put it back to being just a spare?Since it isn''t the rpool, you shouldn''t have to partition the replacement drive. Since you''ve physically replaced the drive, you should just have to do: # zpool replace tank c10t0d0 The pool should resilver, and I think the spare should automatically detach. If not # zpool remove tank c10t6d0 should take care of it. -B -- Brandon High : bhigh at freaks.com
It''s resilvering now - thanks for the help! ----- Original Message ----- From: "Cindy Swearingen" <cindy.swearingen at oracle.com> To: "Chris Dunbar - Earthside, LLC" <cdunbar at earthside.net> Cc: zfs-discuss at opensolaris.org Sent: Friday, July 22, 2011 4:26:00 PM Subject: Re: [zfs-discuss] Replacing failed drive Chris, You don''t need to partition the drive because your pool contains whole disks. The zpool replace command will do any relabeling that is needed. Do the replace first and then detach the spare, if needed. The spare should detach automatically but sometimes it doesn''t. Thanks, Cindy On 07/22/11 14:12, Chris Dunbar - Earthside, LLC wrote:> Cindy, > > Thank you for the reply. I am running Solaris Express 11 (5.11 snv_151a > i86pc i386 i86pc Solaris). > > I have physically replaced the drive, but I have not partitioned it yet. > I know there is a command to copy the layout from one disk to another > and that has worked well for me in the past. I just have to find the > command again. Once that is done, do I need to detach the spare before I > run the replace command or does running the replace command > automatically bump the spare out of service and put it back to being > just a spare? > > Thanks! > Chris > > ------------------------------------------------------------------------ > *From: *"Cindy Swearingen" > *To: *"Chris Dunbar - Earthside, LLC" > *Cc: *zfs-discuss at opensolaris.org > *Sent: *Friday, July 22, 2011 3:57:00 PM > *Subject: *Re: [zfs-discuss] Replacing failed drive > > Hi Chris, > > Which Solaris release is this? > > Depending on the Solaris release, you have a couple of > different options. > > Here''s one: > > 1. Physically replace original failed disk and detach > the spare. > > A. If c10t0d0 was the disk that you physically replaced, > issue this command: > > # zpool replace tank c10t0d0 > # zpool clear tank > > Then check zpool status to see if the spare detached > automatically. > > B. If zpool status says that c10t0d0 is ONLINE but the > spare is still INUSE, just detach the spare: > > # zpool detach tank c10t6d0 > > The next one depends on your Solaris release. > > Thanks, > > Cindy > > > On 07/22/11 12:35, Chris Dunbar - Earthside, LLC wrote: > > Hello, > > > > The bad news is I lost a drive on my production server. The good news > is the spare kicked in and everything kept running as it should. I > stopped by the datacenter last night, shut down the server, and > physically replaced the failed drive. I want to make sure I don''t screw > anything up so would somebody be so kind as to lay out what my next > steps are? Below is a snippet of what zpool status is showing: > > > > tank DEGRADED 0 0 0 > > mirror-0 DEGRADED 0 0 0 > > spare-0 DEGRADED 0 0 0 > > c10t0d0 REMOVED 0 0 0 > > c10t6d0 ONLINE 0 0 0 > > c11t0d0 ONLINE > > > > snip... > > > > spares > > c10t6d0 INUSE currently in use > > > > > > How do to bring the replaced drive back online and get it into the > array? Do I make the new drive the spare or do I bring the new drive > online in the mirror and return the original spare to spare status? Any > advice and/or actual commands would be greatly appreciated! > > > > Thank you, > > Chris Dunbar > > _______________________________________________ > > zfs-discuss mailing list > > zfs-discuss at opensolaris.org > > http://mail.opensolaris.org/mailman/listinfo/zfs-discuss > >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.opensolaris.org/pipermail/zfs-discuss/attachments/20110722/054532aa/attachment.html>