Degraded RAID-Z1 pool and unused disk after expansion - continued

TrueNAS General
,
1

Hi everybody,
as a newbie I am only allowed to post 3 times to a thread (sorry, that’s nonsense), so I have to abandon the old one:

here is what I did after the useful tips from @etorix and @SmallBarky

  • I removed the 1TB cache drive from my dataset
  • I wiped the (in general usable) 4th drive
  • I added it again to start resilvering with it

after another 10hrs of wait time now ALL WORKS.
It was still a bug, I’m sure, but at least I’m up and running now with all 4 raidz1 drives after the expansion.

regards,
Mike

3 Likes
2

Keep us updated. The limit on posts is probably related to the spammer posts we were seeing on this forum this past year.

I don’t know if you get any automated messages from the ‘forum bot’ with a welcome message and such but I ended up working my way through a few forum tutorials and completing tasks for ‘badges’ and upped my rank over a little bit of time.

You could also try reaching out to Admin or Moderator for any related questions to posting limits, etc.

https://forums.truenas.com/about

3

big thanks @SmallBarky , I got a message to be upped a rank now, so probably it won’t happen again :slight_smile:
back to the topic: I’m sorry that I worked around the issue, because I’m sure, this is a bug somehow…
anyway, now, after expanding my zraid1, I was planning to run the zfs-inplace-rebalancing script on github, but this will take forever and I’m not really sure if it is really necessary…

truenas_admin@truenas:~$ sudo zpool list -v
NAME                                       SIZE  ALLOC   FREE  CKPOINT  EXPANDSZ   FRAG    CAP  DEDUP    HEALTH  ALTROOT
RAID                                      58.2T  29.5T  28.7T        -         -     0%    50%  1.00x    ONLINE  /mnt
  raidz1-0                                58.2T  29.5T  28.7T        -         -     0%  50.7%      -    ONLINE
    a83058de-032f-4f81-9c0a-f62bb50cc359  14.6T      -      -        -         -      -      -      -    ONLINE
    bb9f4f4b-9457-4f2b-a398-b9eab36d5e96  14.6T      -      -        -         -      -      -      -    ONLINE
    6fb1e42b-76a5-44a4-b63a-c608e643435a  14.6T      -      -        -         -      -      -      -    ONLINE
    ca7fea98-7ecb-45de-9b1d-efd652dbc15d  14.6T      -      -        -         -      -      -      -    ONLINE
4

Not sure on re-balancing. I thought that was part of the expansion process. I haven’t read up on what exactly it does.

5

I had this same question yesterday and concluded that rebalancing is not a big enough issue to worry about. Refer here:

6

There are three issues with Raidz Expansion

  1. it takes a relatively long time… this is solved by waiting patiently. You can restart etc safely.
  2. old data has the old storage efficiency. So if your raidz array had a storage efficiency of 2/3 (ie 66% or 2 data disks + 1 parity disk) before expansion the old data will still have that storage efficiency, although new data would have 3/4, or 75%. Rebalancing or otherwise rewriting the old data turns it into new data and will mean it’s stored more efficiently.
  3. Zfs Raidz Space Accounting is whacked. Which there may be a future fix. One day. Maybe.
1 Like