zpool replace leaves pool degraded

[Bobo Moreno]

hi,
I've been struggling with a failing disk in my zpool, FreeNAS 9.10

One of the drives was being warned as having many bad sectors so I decided to replace it. Had tremendous problems getting 'zpool replace' to recognise the new drive (it was in ada3, but it wouldn't accept that as parameter). Only thing I could figure out was to add the new drive as a 'spare' to the zpool and then used "zpool replace [poolname] [old device id] [spare device id]"

This worked and resilvered the new drive. However, once the resilver completed the pool remains in degraded state and seems to want the old drive back.

How do I convince it to 'forget' the old drive and accept the new one permanently and let the 3 drives play happy pool together??
Sorry if this is obvious, I've been googling all day.

Thanks for any input.

pool: ZFS_NAS

state: DEGRADED

status: One or more devices could not be opened. Sufficient replicas exist for

the pool to continue functioning in a degraded state.

action: Attach the missing device and online it using 'zpool online'.

see: http://illumos.org/msg/ZFS-8000-2Q

scan: resilvered 989G in 4h42m with 0 errors on Mon May 2 19:45:33 2016

config:


NAME STATE READ WRITE CKSUM

ZFS_NAS DEGRADED 0 0 0

raidz1-0 DEGRADED 0 0 0

spare-0 DEGRADED 0 0 0

12082773611957310038 UNAVAIL 0 0 0 was /dev/gptid/1418d56c-431b-11e4-b9f7-28924a2f106f

gptid/503d6d1c-106e-11e6-a169-28924a2f106f ONLINE 0 0 0

gptid/1608e28a-431b-11e4-b9f7-28924a2f106f ONLINE 0 0 0

gptid/1699dab6-431b-11e4-b9f7-28924a2f106f ONLINE 0 0 0

spares

16673430511205791764 INUSE was /dev/gptid/503d6d1c-106e-11e6-a169-28924a2f106f


errors: No known data errors​

 

[depasseg]

I'm guessing you did this at the command line, instead of the GUI. Is that correct?

If so, a reboot should fix it. The GUI has no idea what you've done behind the scenes.

In the future this might be a useful reference:
http://doc.freenas.org/9.3/freenas_storage.html#replacing-a-failed-drive

 

[Bobo Moreno]

Hi,
thanks, I had to do it from the command line as GUI was not showing anything at all (no drives, no pool, etc.). I later discovered that this is due to a Chrome bug as it works in other browsers, but by then too late.

I have rebooted multiple times and the output shown above is same on CLI and GUI.

Would really appreciate any advice that could fix this.

 

[depasseg]

If a reboot didn't fix it, then the only other way I know of, would be to export and then import the pool using the GUI for both steps.

 

[Robert Smith]

I think you just need to detach the unavailable drive, and then the already reslivered spare should show up in its place. I never tried using spares with FreeNAS, though, as I have been warned that FreeNAS does not support them.

Do not remove an INUSE spare.