New NAS Help

[Bruce A]

I grew tired of my old QNap home Nas throwing and error on bay-2 constantly so put together a nas with an older motherboard.

I have 6 3Tb drives in a pool with ZFS1 and a hot spare. One of the drive errored out, the hot spare took over. However, when I went to replace the drive the new drive appears to have attached ok, the system completed the re-sliver however the pool now shows as unhealthy and the hotspare still shows as part of the pool, what have I done wrong here?

NAS is running TrueNAS-12.0-U1 with 16Gb Ram, 6 WD 3Tb HDD and an Intel 100Gb SSD for boot on an i5-2400t.

I don't see any errors on the drives, but then I may be looking in the wrong place.

 

[Samuel Tai]

Please provide the output of zpool status Pool1. This will have a better description of the errors seen.

the hotspare still shows as part of the pool, what have I done wrong here?

View attachment 44802

You haven't done anything wrong. This is just how spares display. It's showing that ada6p2 is acting as a spare for ada2p2.

 

[Bruce A]

Looks like this.

 

[Samuel Tai]

No, try zpool status -v Pool1.

 

[Bruce A]

Looks like an error in a specific file, Do i just go ahead and delete this to clear up the error?

 

[Samuel Tai]

Yes, and then scrub your pool. You may need to scrub a couple of times to get rid of the error. Then you can run the disk replace procedure.

 

[Bruce A]

Thank you, I did replace the failed drive already, should I remove and restart the replace procedure?

 

[Samuel Tai]

No, just click the 3 dots by ada2p2, and select replace.

 

[Bruce A]

Thank you for your help, this cleared the error perfectly.

The Hot spare drive is still showing up in the ADA2p2 space though, even after replacing the original bad drive. If I click on replace again it doesn't give me any disk options, any ideas? I do have a spare drive I could replace it again?

 

[Dirk]

Thank you for your help, this cleared the error perfectly.

The Hot spare drive is still showing up in the ADA2p2 space though, even after replacing the original bad drive. If I click on replace again it doesn't give me any disk options, any ideas? I do have a spare drive I could replace it again?

Did you get this resolved? I'm having the same issue with a duplicate spare.

 

[sretalla]

I have 6 3Tb drives in a pool with ZFS1 and a hot spare

You need to return it to the spares from CLI with a zpool detach command (https://www.truenas.com/community/t...nnot-free-up-the-hot-spare.90059/#post-623314), but I would generally advise against using a "spare" approach, particularly with RAIDZ1.

If any of the drives in the pool fail, the spare will start resilvering, but that spare has been spinning the whole time (so having it there as a spare isn't really winning you anything at all in terms of drive life or power consumption), so could have just been a member of a RAIDZ2 and already had the resilvering up to date as a regular pool member.

 

[Dirk]

You need to return it to the spares from CLI with a zpool detach command (https://www.truenas.com/community/t...nnot-free-up-the-hot-spare.90059/#post-623314), but I would generally advise against using a "spare" approach, particularly with RAIDZ1.

If any of the drives in the pool fail, the spare will start resilvering, but that spare has been spinning the whole time (so having it there as a spare isn't really winning you anything at all in terms of drive life or power consumption), so could have just been a member of a RAIDZ2 and already had the resilvering up to date as a regular pool member.

Thanks sretalla. Further research revealed the command line to do and it's back to normal. The setup is a Z2 with 5 data and the hot swap. The spare is set to spin down and sleeps after 5 minutes

 

[sretalla]

The spare is set to spin down and sleeps after 5 minutes

As long as you've convinced yourself that it is actually spinning down, that's OK. Maybe check it with the SMART stats.

 

[Dirk]

As long as you've convinced yourself that it is actually spinning down, that's OK. Maybe check it with the SMART stats.

I was skeptical as well, but found the below command that confirmed it is spinning down. I suppose the real test would/will be to see it wakes up and joins the pool.

camcontrol cmd ada6 -a "E5 00 00 00 00 00 00 00 00 00 00 00" -r-