DegradedPool - Revision history

86.103.157.37 at 19:28, 13 May 2014

2014-05-13T19:28:39Z

Ilovezfs at 19:43, 1 April 2014

2014-04-01T19:43:46Z

84.114.59.37 at 19:18, 1 April 2014

2014-04-01T19:18:05Z

84.114.59.37 at 19:16, 1 April 2014

2014-04-01T19:16:30Z

84.114.59.37: replace drive in degraded pool

2014-04-01T19:15:23Z

replace drive in degraded pool

New page

In case you have a degraded pool due to corrupted data on one drive, zpool status will show something like this:

<syntaxhighlight lang="bash">sudo zpool status
pool: tank
id: 4167705491194610230
state: DEGRADED
status: One or more devices contains corrupted data.
action: The pool can be imported despite missing or damaged devices. The
fault tolerance of the pool may be compromised if imported.
see: http://zfsonlinux.org/msg/ZFS-8000-4J
config:

tank DEGRADED
fs DEGRADED
disk1s2 ONLINE
disk2s2 ONLINE
2735026605872377559 FAULTED corrupted data
disk4s2 ONLINE
</syntaxhighlight>

Now, look into DiskUtility and find the disk number of the disk which is missing. (In our case disk3).
If you are certain that the corrupted data was not due to a bad disk, (for example if just the labels were messed up), we would like to delete the partition table of that disk and afterwards format the drive.

If the disk is faulty, take a new disk and skip this step, going directly to the format.

<syntaxhighlight lang="bash">
sudo gpt destroy disk3
</syntaxhighlight>

Now, format in DiskUtility as one partition, using GUID as hfs+ and look up the new disk partition id in the lower part of the DiskUtility-window. (in our case disk3s2)
Replace the now missing drive in the pool with the newly formatted disk:

<syntaxhighlight lang="bash">
sudo zpool replace tank 2735026605872377559 /dev/disk3s2
</syntaxhighlight>

The number you see is the one which you saw above at the position of the faulted drive in <syntaxhighlight lang="bash">sudo zpool status</syntaxhighlight>.

The system now starts resilvering the drive. This takes a long time. With <syntaxhighlight lang="bash">sudo zpool status</syntaxhighlight> you can see the progress.

@@ Line 19: / Line 19: @@
              disk2                ONLINE       0     0     0
              disk1                ONLINE       0     0     0
-             2735026605872377559  FAULTED      0     0     0  was /dev/disk1
+             2735026605872377559  FAULTED      0     0     0  was /dev/disk3
              disk4                ONLINE       0     0     0

@@ Line 1: / Line 1: @@
 In case you have a degraded pool due to corrupted data on one drive, zpool status will show something like this:
-<syntaxhighlight lang="bash">sudo zpool status
+<syntaxhighlight lang="bash">$ sudo zpool status</syntaxhighlight>
- pool: tank
-     id: 4167705491194610230
+<syntaxhighlight lang="text">
-  state: DEGRADED
+  pool: tank
- status: One or more devices contains corrupted data.
+ state: DEGRADED
- action: The pool can be imported despite missing or damaged devices.  The
+status: One or more devices could not be used because the label is missing or
-         fault tolerance of the pool may be compromised if imported.
+        invalid.  Sufficient replicas exist for the pool to continue
     see: http://zfsonlinux.org/msg/ZFS-8000-4J
- config:
+  scan: none requested
-         tank                                      DEGRADED
+        NAME                     STATE     READ WRITE CKSUM
-           fs                                        DEGRADED
+         tank                     DEGRADED     0     0     0
-             disk1s2                             ONLINE
+           raidz1-0               DEGRADED     0     0     0
-             disk2s2                             ONLINE
+             disk2                ONLINE       0     0     0
-             2735026605872377559   FAULTED  corrupted data
+             disk1                ONLINE       0     0     0
-             disk4s2                             ONLINE
+             2735026605872377559  FAULTED      0     0     0  was /dev/disk1
 </syntaxhighlight>
-Now, look into DiskUtility and find the disk number of the disk which is missing. (In our case disk3).
+Now, look into Disk Utility and find the disk number of the disk which is missing. (In our case disk3).
-If you are certain that the corrupted data was not due to a bad disk, (for example if just the labels were messed up), we would like to delete the partition table of that disk and afterwards format the drive.
+If you are certain that the corrupted data was not due to a bad disk, (for example if just the labels were messed up), we can reuse the same disk. First, delete the partition table of that disk.
-−
-If the disk is faulty, take a new disk and skip this step, going directly to the format.
 <syntaxhighlight lang="bash">
@@ Line 28: / Line 32: @@
 </syntaxhighlight>
-Now, format in DiskUtility as one partition, using GUID partition table, as hfs+ and look up the new disk partition id in the lower part of the DiskUtility-window. (in our case disk3s2)
+Replace the missing device in the pool with this disk.
-Replace the now missing drive in the pool with the newly formatted disk:
 <syntaxhighlight lang="bash">
-sudo zpool replace tank 2735026605872377559 /dev/disk3s2
+sudo zpool replace -f tank 2735026605872377559 /dev/disk3
 </syntaxhighlight>

@@ Line 35: / Line 35: @@
 </syntaxhighlight>
-The number you see is the one which you saw above at the position of the faulted drive in <syntaxhighlight lang="bash">sudo zpool status</syntaxhighlight>.
+The number you see is the one which you saw above at the position of the faulted drive in '''sudo zpool status'''.
-The system now starts resilvering the drive. This takes a long time. With <syntaxhighlight lang="bash">sudo zpool status</syntaxhighlight> you can see the progress.
+The system now starts resilvering the drive. This takes a long time. With '''sudo zpool status''' you can see the progress.