[SOLVED] "read-only file system" after crash of cache pool

[JorgeB]

Yes.

[PzrrL]

I can't wait and just started without checking the box, will that be alright?

[PzrrL]

Result:

 

UUID:             hidden
Scrub started:    Thu Oct 13 00:57:18 2022
Status:           finished
Duration:         0:09:33
Total to scrub:   412.41GiB
Rate:             736.64MiB/s
Error summary:    no errors found

 

[JorgeB]

Looks good for the moment, and now if one device fails pool should continue working, but of course should be fixed ASAP, also see here for better pool monitoring, you'll need to reset the current errors as explained there.

[PzrrL]

@JorgeB million thanksssss! I have already setup the monitor following the post provided. May you also provide guidance on swapping out the faulty cache drive? Besides, do you recommend do the upgrade of unraid first or swapping first? Thanks.

[JorgeB]

Upgrade first since that functionally is broken on v6.9, then see here:

https://forums.unraid.net/topic/46802-faq-for-unraid-v6/?do=findComment&comment=480419

 

[PzrrL]

@JorgeB Unfortunately the monitor is raising errors not long after the scrub. Do I need to re-do all the steps in order to make sure the cache drives are in sync before swapping?

 

Regarding the link you sent, can I do the following as I don't have enough ports:

1. Stop the array and shutdown

2. Swap the faulty ssd

3. Power Up

4. Assign the new device in the faulty's SSD slot?

5. Start the array

6. Wait for cache copying to complete

 

Is this procedure okay? Thanks!

[JorgeB]

5 hours ago, PzrrL said:

Do I need to re-do all the steps in order to make sure the cache drives are in sync before swapping?

Depends on if the errors are from the device you're going to remove, best to post new diags.

[PzrrL]

#  btrfs dev stats /mnt/cache
[/dev/sdi1].write_io_errs    475032
[/dev/sdi1].read_io_errs     3675
[/dev/sdi1].flush_io_errs    55531
[/dev/sdi1].corruption_errs  0
[/dev/sdi1].generation_errs  0
[/dev/sdj1].write_io_errs    0
[/dev/sdj1].read_io_errs     0
[/dev/sdj1].flush_io_errs    0
[/dev/sdj1].corruption_errs  0
[/dev/sdj1].generation_errs  0

uncrate-diagnostics-20221013-2223.zip

 

Please have a look thanks!

[JorgeB]

It's the Intel SSD again, if that's the one you were going to replace you can go ahead, and the steps you mentioned above are correct.

[PzrrL]

Thanks I just swapped the disk. Should I actually upgrade first? Or it will be fine if I do the swapping then the upgrade?

[JorgeB]

Like mentioned before you must upgrade first or it won't work.

[PzrrL]

Sorry I am a bit confused on the steps right now.

 

I should

1. start the array with both cache drive (the old one)
2. perform an upgrade
3. Confirm upgrade successful
4. Swap cache drive as the steps I mentioned above
5. rebuild the cache pool

 

Is this correct?

[JorgeB]

You can do the upgrade now and instead of rebooting power off and replace the SSD, then just check that it's on the new release after boot.

[PzrrL]

Thanks @JorgeB I did an upgrade and see I am on 6.11.1 now. I also replaced the SSD and built the pool without problem. Is there any final check needed? Thanks!

uncrate-diagnostics-20221014-2142.zip

Hopefully this will be the final diags.

[JorgeB]

Pool looks good so far.

[PzrrL]

Great thanks a million @JorgeB you have been really patient and helpful!!! Speechless thanks going through this with me!!!