Hello there,
I am aware that we are entering into an intense two days event :-) but I would
ask your opinions regarding this issue.
We recently have got a faulted pool most probably due to an HA switch where one
host has tried to import it when the other had it still imported.
This is how looks like the "
zpool
import" command (on both hosts)
bash-4.4$ pfexec zpool import
pool: sp0
id: 3741442844850542262
state: FAULTED
status: The pool metadata is corrupted.
action: The pool cannot be imported due to damaged devices or data.
The pool may be active on another system, but can be imported using
the '-f' flag.
see: http://illumos.org/msg/ZFS-8000-72
config:
sp0 FAULTED corrupted data
raidz2-0 ONLINE
c1t5000CCA2521ACF35d0 ONLINE
c1t5000CCA25218C2A1d0 ONLINE
c1t5000CCA2521A130Dd0 ONLINE
c1t5000CCA2521B9FD1d0 ONLINE
c1t5000CCA2521ACFFDd0 ONLINE
c1t5000CCA251B0CC7Dd0 ONLINE
c1t5000CCA2515ED459d0 ONLINE
raidz2-1 ONLINE
c1t5000CCA2521BA3E9d0 ONLINE
c1t5000CCA25143ADD9d0 ONLINE
c1t5000CCA252194CDDd0 ONLINE
c1t5000CCA2521939C9d0 ONLINE
c1t5000CCA2521ACF65d0 ONLINE
c1t5000CCA2521B7E4Dd0 ONLINE
c1t5000CCA2521ACE61d0 ONLINE
raidz2-2 ONLINE
c1t5000CCA2661BCE42d0 ONLINE
c1t5000CCA26615E36Ad0 ONLINE
c1t5000CCA2515E1A56d0 ONLINE
c1t5000CCA2662261A6d0 ONLINE
c1t5000CCA2660DE6A6d0 ONLINE
c1t5000CCA26618BA9Ed0 ONLINE
c1t5000CCA25193AC22d0 ONLINE
raidz2-3 ONLINE
c1t5000CCA25130E6DEd0 ONLINE
c1t5000CCA2660E6E7Ed0 ONLINE
c1t5000CCA26622EC76d0 ONLINE
c1t5000CCA2514BC8CAd0 ONLINE
c1t5000CCA2660FBC1Ed0 ONLINE
c1t5000CCA251B0DC66d0 ONLINE
c1t5000CCA25189BE3Ad0 ONLINE
raidz2-4 ONLINE
c1t5000CCA26620B24Ed0 ONLINE
c1t5000CCA2661C891Ed0 ONLINE
c1t5000CCA2661C4036d0 ONLINE
c1t5000CCA26619EA4Ed0 ONLINE
c1t5000CCA266219F16d0 ONLINE
c1t5000CCA2515FF7C6d0 ONLINE
c1t5000CCA2513E91CEd0 ONLINE
raidz2-5 ONLINE
c1t5000CCA25198D01Ed0 ONLINE
c1t5000CCA2513315B2d0 ONLINE
c1t5000CCA2513291D2d0 ONLINE
c1t5000CCA26610E5A6d0 ONLINE
c1t5000CCA2513477C6d0 ONLINE
c1t5000CCA2661EFA82d0 ONLINE
c1t5000CCA2518FEEF2d0 ONLINE
raidz2-8 ONLINE
c1t5000CCA25219009Ad0 ONLINE
c1t5000CCA2521924DEd0 ONLINE
c1t5000CCA26629ADC6d0 ONLINE
c1t5000CCA251820F72d0 ONLINE
c1t5000CCA25218A40Ad0 ONLINE
c1t5000CCA252194046d0 ONLINE
c1t5000CCA252191AA2d0 ONLINE
raidz2-9 ONLINE
c1t5000CCA251B13DDAd0 ONLINE
c1t5000CCA266260492d0 ONLINE
c1t5000CCA2521B810Ad0 ONLINE
c1t5000CCA2521B36D2d0 ONLINE
c1t5000CCA2662DF89Ed0 ONLINE
c1t5000CCA25199F0E6d0 ONLINE
c1t5000CCA252191A0Ed0 ONLINE
logs
c1t5000C500AD7048D6d0 ONLINE
c1t5000C500AD7048E6d0 ONLINE
This pool should have also a cache and a hot-spare group which are not listed
anymore. We have checked the drives and they appear to be available to the
system.
We have tried to import the pool with these commands
bash-4.4$ pfexec zpool import -o readonly=on -f sp0
cannot import 'sp0': I/O error
Destroy and re-create the pool from
a backup source.
bash-4.4$ pfexec zpool import -o readonly=on -nF sp0
this import command simply returns without any message but the pool is not
imported.
bash-4.4$ pfexec zpool import -o readonly=on -FfX -T 23486835 sp0
cannot import 'sp0': one or more devices is currently unavailable
bash-4.4$ pfexec zpool import -o readonly=on -d /devices -FfX -T 23486835 sp0
cannot import 'sp0': no such pool available
Would it be possible to let
zfs
to skip the pool topology check or change it in a way that the raid groups
will be in the right sequence (group 8 and 9 should be 6 and 7
).
Best, Luca
------------------------------------------
openzfs: openzfs-developer
Permalink:
https://openzfs.topicbox.com/groups/developer/T4f4949cfdd8ef028-Mdeba902ed6e7fb3afd6b1b08
Delivery options: https://openzfs.topicbox.com/groups/developer/subscription
