[zfs-discuss] Replacing disk in RAIDZ2 pool with autoreplace=on
mabi at protonmail.ch
Mon Nov 12 08:53:02 EST 2018
I am planning to replace a single disk on my RAIDZ2 12-disks ZFS pool. This pool has the autoreplace=on attribute set. This means that I can simply take the disk out and replace it with another disk in the same slot without having to enter a single command on the server and the resilver will happen automatically, correct?
I was also wondering because of the ashift=12 parameter which I use on this pool. Will the new replacement disk immediately inherit this ashift=12? Because based on the zpool man page this seems not to be the case when manually running a "zpool replace".
For the sake of completeness I am replacing this 6 TB SAS disk as it shows some signs of failure after this monthly pool scrub as you can see below.
sd 0:0:0:0: [sda] tag#2 FAILED Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE
sd 0:0:0:0: [sda] tag#2 Sense Key : Medium Error [current] [descriptor]
sd 0:0:0:0: [sda] tag#2 Add. Sense: Unrecovered read error
sd 0:0:0:0: [sda] tag#2 CDB: Read(16) 88 00 00 00 00 01 16 20 3b 38 00 00 00 f8 00 00
blk_update_request: critical medium error, dev sda, sector 4666178576
relevant disk from "zpool status":
status: One or more devices has experienced an unrecoverable error. An
attempt was made to correct the error. Applications are unaffected.
action: Determine if the device needs to be replaced, and clear the errors
using 'zpool clear' or replace the device with 'zpool replace'.
scan: scrub repaired 108K in 21h28m with 0 errors on Sun Nov 11 21:52:40 2018
NAME STATE READ WRITE CKSUM
data ONLINE 0 0 0
raidz2-0 ONLINE 0 0 0
sda ONLINE 10 0 0
and finally relevant output from "smartctl -a /dev/sda" after a longtest:
SMART Self-test log
Num Test Status segment LifeTime LBA_first_err [SK ASC ASQ]
Description number (hours)
# 1 Background long Failed in segment --> - 25535 4666178576 [0x3 0x11 0x0]
For those who are curious, this is a nearly 4 years old Seagate enterprise ST6000NM0034 which has a warranty of 5 years.
More information about the zfs-discuss