Issues with new drive in cache pool

[obuwunkunubi]

Hello,

so today I decided to add a drive to my cache pool. It was going normally until the drive temperature randomly stopped showing. But the balance seemed to keep on going. The problem is that after it finished the mover wasn't working. So I decided to shut down the system and try another SATA cable. The temperature and smart details started showing again after starting up again but I decided to run a btrfs scrub to check for errors. It was correcting what seemed like every single sector until it started showing I/O errors (for the new disk) in the logs and subsequently uncorrectable sectors, so I stopped the scrub. It seems that I can still access the files on the pool since the first drive seems to be working normally. The new drive has since stopped showing smart info and temperature again. Is there a way to fix these issues? Is it a problem with the drive, since it doesn't seem to be the SATA cable? In that case can I remove the new drive and only keep the original while I wait for a replacement?

Otherwise I have made an appdata backup before doing this and I also backed up some of the more important files before doing this. But I would still like to keep everything if possible (even though they are mostly torrents which I can re-download)

 

I have attached my diagnostics file below. (Sadly I forgot to save one before shutting the system down to switch the SATA cable, so this is just the new one)

serverina-diagnostics-20220818-2150.zip

Edited August 18, 2022 by obuwunkunubi

[JorgeB]

Looks more like a connection/power problem, try a different SATA port, also make sure you're using a different power cable if you only replaced the SATA cable.

[obuwunkunubi]

1 hour ago, JorgeB said:

try a different SATA port, also make sure you're using a different power cable

 

Thanks for the suggestion, I have tried both now. Things such as smart info are still showing for now but my log is still filled with errors of the following two types (in the code box below (sdc is the new drive that has been giving me issues)) whenever I try to access a file that is stored on the cache. The file still opens and plays normally otherwise.

 

Aug 19 11:23:06 Serverina kernel: BTRFS error (device sdb1): bad tree block start, want 1271081533440 have 0
Aug 19 11:23:06 Serverina kernel: BTRFS error (device sdb1): bad tree block start, want 1271081549824 have 0

Aug 19 11:23:12 Serverina kernel: BTRFS info (device sdb1): read error corrected: ino 0 off 1270182952960 (dev /dev/sdc1 sector 944117472)
Aug 19 11:23:12 Serverina kernel: BTRFS info (device sdb1): read error corrected: ino 0 off 1270182957056 (dev /dev/sdc1 sector 944117480)

 

So should I try to run another scrub to repair the errors?

 

I have also noticed that it says "No balance found on '/mnt/cache'". Is that also an issue? Should I be performing another balance?

[JorgeB]

1 hour ago, obuwunkunubi said:

So should I try to run another scrub to repair the errors?

Yes.

 

[obuwunkunubi]

1 hour ago, JorgeB said:

Yes.

 

Thanks, that seems to have worked! There were no uncorrectable errors and there were no I/O errors in the logs. There are also no errors in the logs when trying to open a file.

 

Do I have to take any further steps regarding my cache pool or can I use it normally now?

It still says "No balance found on '/mnt/cache'", is that an issue or is it supposed to be like that?

[JorgeB]

1 hour ago, obuwunkunubi said:

is it supposed to be like that?

This, unless a balance is running, pool is OK for now, recommend taking a look here for better pool monitoring.

[obuwunkunubi]

7 minutes ago, JorgeB said:

unless a balance is running, pool is OK for now

 Since the "Balance" button isn't greyed out and I am not seeing any balance progress I'm assuming that it isn't running.

 

Thanks for the help and suggestions! I'll take a look at the monitoring script.