0

I maintain a Debian server with a ZFS storage pool (RAID Z3). Recently ZFS reported that two disks had failed simultaneously:

ZFS has detected that a device was removed.

 impact: Fault tolerance of the pool may be compromised.
    eid: 138
  class: statechange
  state: REMOVED
   host: serres-west-wing
   time: 2021-04-30 01:30:15+0300
  vpath: /dev/disk/by-vdev/d0-part1
  vguid: 0x6622AF6B1929E199
   pool: 0x0964CF6A3748D7A9
ZFS has detected that a device was removed.

 impact: Fault tolerance of the pool may be compromised.
    eid: 140
  class: statechange
  state: REMOVED
   host: serres-west-wing
   time: 2021-04-30 01:30:15+0300
  vpath: /dev/disk/by-vdev/d1-part1
  vguid: 0xD48BA6B066788199
   pool: 0x0964CF6A3748D7A9

The hot spare was activated and a resilvering started immediately after these messages were generated. The status of the pool after resilvering is the following:

ZFS has finished a resilver:

   eid: 167
 class: resilver_finish
  host: serres-west-wing
  time: 2021-04-30 02:15:03+0300
  pool: datapool
 state: ONLINE
  scan: resilvered 132G in 00:44:41 with 0 errors on Fri Apr 30 02:15:03 2021
config:

        NAME               STATE     READ WRITE CKSUM
        datapool           ONLINE       0     0     0
          raidz2-0         ONLINE       0     0     0
            spare-0        ONLINE       0     0     0
              d0-part1     ONLINE       0     0     0
              hs-d0-part1  ONLINE       0     0     0
            d1-part1       ONLINE       0     0     0
            d2-part1       ONLINE       0     0     0
            d3-part1       ONLINE       0     0     0
            d4-part1       ONLINE       0     0     0
        logs
          mirror-1         ONLINE       0     0     0
            zil-d0-part1   ONLINE       0     0     0
            zil-d1-part1   ONLINE       0     0     0
        cache
          l2arc-d0-part2   ONLINE       0     0     0
          l2arc-d1-part2   ONLINE       0     0     0
        spares
          hs-d0-part1      INUSE     currently in use

errors: No known data errors

The disks d0-part1 and d1-part1 seem to be connected and work fine.

Was this an error due to factors unrelated to disk degredation? It seems unlikely that two working disks failed at once. Is it safe to deactivate the hot spare?

1 Answer 1

0

It seems that disk disconnections were caused by a power supply problem. After upgrading the UPS for the machine I haven't encountered any issues. I deactivated the hot spare

zpool detach datapool hs-d0-part1

and then I resilvered the pool

zpool scrud datapool

to restore the pool to its original state.

1
  • Please don't forget to check back and accept your answer after 48 hours. Commented Oct 2, 2021 at 17:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .