Thank you

So - following your guide I've successfully changed one of the pool ("Cache") devices to a new one. But today I've noticed that my log file is 87% full... 
Have I changed the wrong "Cache" pool device? 

the-beast-diagnostics-20240707-1027.zip

Before there were errors on both pool devices, so possibly both are bad, or there's some power/connection or even compatibility issue with them, you can start by replacing the cables first, both power and SATA, and if the issues continue with just that device I would replace it also.

Before upgrade power cables were just fine. Although I've already changed SATA cables, I could say the same about those.

I think I'll just get another new SSD, maybe these will work in my gaming rig.
What should be the sequence to replace another SSD?

1. Scrub
2. BTRFS dev stats -z
3. Scrub again
4. Ensure BTRFS dev stats shows zeros
5. Continue with replacement instructions

Is that correct?

Yes, that should be the safest.

What should I do with these uncorrectable errors?

Look in the syslog for the list of corrupt files and delete them/restore from a backup.

Can't find any, just errors...

I did back up my /mnt/cache before SSD swap. I can live with couple of days loss, but what should be the procedure?

the-beast-diagnostics-20240710-1643.zip

P.S. As far as I understand, now only newly installed SSD (sdg) spits out errors? So it seams problem might be elsewhere (in power cord or the PSU itself?)

42 minutes ago, JunkoZane said:

So it seams problem might be elsewhere (in power cord or the PSU itself?)

Very possibly.

Hi there, turns out photoprism container was corrupt. Those error would start right after I'd start and use it.

I'm glad my problems stopped, it will be a good opportunity to try out immich

 

Thank you for your help!