Strange behaviour disc status

[xbmcgotham]

Hi,

Hope someone can point me to the right direction. For some reason I have a disc (da21) that shows UNAVAIL and whatever I try, it wont go online. I also did a SMART test and this returns a succesful test. Additionally, this same disc shows up twice in both spare location and there it shows the status ONLINE.

I have attached the screenshots of the pool disc status.


Kind regards

 

[sretalla]

It's a spare that's currently in use...

You need to decide what you want to... either keep it in the pool or put it back as a spare.

To put it back to spare, you should be able to zpool detach RAIDZ3 /dev/da21p2 (or maybe you will need to look up the gptid instead with glabel status and use that instead).

 

[xbmcgotham]

It's a spare that's currently in use...

You need to decide what you want to... either keep it in the pool or put it back as a spare.

To put it back to spare, you should be able to zpool detach RAIDZ3 /dev/da21p2 (or maybe you will need to look up the gptid instead with glabel status and use that instead).

Thanks for your quick response!

Do I need to do this for both da21p2 and da20p2? As I have 2 spares in my pool.

 

[sretalla]

Do I need to do this for both da21p2 and da20p2? As I have 2 spares in my pool.

I would recommend handling it somehow... spares shouldn't stay there forever as they are currently (or they won't be any good to you as spares and shouldn't be there at all anyway).

 

[xbmcgotham]

Hi again @sretalla,

I tried the above code and only changed the pool name to reflect my own. However I am getting the following message when executing:

"cannot detach /dev/da21p2: no such device in pool"

I can try to run the same command for disc da20p2, but want to make sure that I can, without doing something that can break my pool. Please advise.

 

[xbmcgotham]

Hi @sretalla hope you have a moment to check my last post.

 

[sretalla]

or maybe you will need to look up the gptid instead with glabel status and use that instead

zpool status may help with that

 

[xbmcgotham]

zpool status may help with that

Thanks for getting back to me.

The Zpool status confuses me, as I am not able to clearly identify which of the discs shown is the da21p2. In the screenshots of the first post you see that there are two spare discs with status UNAVAIL and ONLINE. But Zpool status below shows two discs with INUSE and AVAIL. May I assume that for both situations the order of the discs is the same? So da21p2 (UNAVAIL) is INUSE?

And would this be the command to try when using gptid:

zpool detach RAIDZ3 gptid/xxxxx-xxx-xxxx

 

[sretalla]

would this be the command to try when using gptid:

zpool detach RAIDZ3 gptid/xxxxx-xxx-xxxx

Yes.

May I assume that for both situations the order of the discs is the same? So da21p2 (UNAVAIL) is INUSE?

Don't assume, use glabel status to link the gptid with the partition/disk identifier

 

[xbmcgotham]

Yes.


Don't assume, use glabel status to link the gptid with the partition/disk identifier

Thanks @sretalla for clearing this up and the tip.

I found the correct gptid that belongs to the da21p2 disc and ran the following command:

zpool detach RAIDZ3 gptid/9dd03cb8-211e-11eb-9b98-002590f392a6

This looked to work as now all of the discs have the status ONLINE. But the Pool status still gives me a message saying: ONLINE (Unhealthy) on the "Pools" page.

There is no indication of any resilver process running either.

My question now is:
1. Why is the status still on unhealthy? Is this because all the discs are online and the two spares should be offline?
2. Should da21p2 not be offline? Or do I need to follow another procedure?

Sorry for the confusion.

Thanks

 

[sretalla]

Why is the status still on unhealthy? Is this because all the discs are online and the two spares should be offline?

Let's see zpool status -v RAIDZ3

 

[xbmcgotham]

Let's see zpool status -v RAIDZ3

please find attached the screenshots. Its interesting to see that here it says both spares are AVAIL but still pool is shown unhealthy.

Thanks again for helping!

 

[sretalla]

if your pool name is pool1, the command you ran earlier was wrong...

zpool detach RAIDZ3 gptid/9dd03cb8-211e-11eb-9b98-002590f392a6 shouldn't have worked... should have been zpool detach pool1 gptid/9dd03cb8-211e-11eb-9b98-002590f392a6

It seems your spare is still attached so you should run zpool detach pool1 gptid/9dd03cb8-211e-11eb-9b98-002590f392a6 and zpool clear pool1 assuming you're happy the disks are all now in good shape and you're ready to go back to monitoring things as normal.

 

[xbmcgotham]

if your pool name is pool1, the command you ran earlier was wrong...

zpool detach RAIDZ3 gptid/9dd03cb8-211e-11eb-9b98-002590f392a6 shouldn't have worked... should have been zpool detach pool1 gptid/9dd03cb8-211e-11eb-9b98-002590f392a6

It seems your spare is still attached so you should run zpool detach pool1 gptid/9dd03cb8-211e-11eb-9b98-002590f392a6 and zpool clear pool1 assuming you're happy the disks are all now in good shape and you're ready to go back to monitoring things as normal.

Hi, Sorry I did change the pool name, but had forgot to change it in my response below.

See response when running it again attached.

So I guess all is fine for now, but I just have to reset the pool errors by running the

zpool clear pool1

command?

.

 

[sretalla]

OK, see if the clear sorts it out.

 

[xbmcgotham]

OK, see if the clear sorts it out.

Great so I also did the same command for da20p2 disc. after this the pool status changed to healthy again!

My last question would be how to properly change discs when a disc become faulty again.

Do I understand correctly that, when a disc needs replacement, TrueNas will automatically take one of the spares that is online so in my case da20p2 or da21p2?

And what correct steps would I need to follow after this to replace the faulty disc?