Both my parity disks keep falling into an error state simultaneously. This has happened 3-4 times now and it feels like it's getting more frequent.

 

I'm thinking it may be that their both plugged directly into the motherboard and those SATA ports are failing. The server is at my Mom's house across the continent so I'll walk her through making sure all the cables are secure and checking where the parity disks are plugged into.

 

In the meantime I really have no idea how to interpret this diagnostics file like at all. If anyone can put me on the right path to getting this fixed it would be hugely appreciated.

closet-diagnostics-20230830-2115.zip

I just confirmed that both parity disks are plugged directly into the motherboard (attached image) instead of the SATA cards. I'm guessing those SATA ports on the motherboard are failing which explains why both drives go out of sync simultaneously.

 

I really should have used HBAs to begin with instead of these cheap-o SATA expansion cards. I just purchased a single Dell H310 from theartofserver on ebay and will be swapping out one of the SATA expansion cards for it so I don't have to use the motherboard SATA ports.

Diags are full of disk related errors and already rotated, so the initial problem is not visible, reboot and post new diags after array start.

Crap. Well I went ahead and unmounted the parity disks at this point. It would take a week or so to rebuild the parity disks and wait for them to come out of sync again so I feel like it's best to just go unprotected till the new HBAs come in next Saturday.

 

I'm having a problem with the Ubuntu VMs now though and I'm wondering if this is an indication of a bigger overall problem with the system. They work for a few days then one day I try to log into them and get this the attached error.

 

Between my surface level understanding of Unraid and searching the errors in the attached error I think whats happening is the VM doesn't see the mounted disks anymore but the only thing I changed was removing the parity disks from the array.

 

Questions:

1. Does the error long 'full of disk related errors' indicate inevitable system failure if nothing is done?
2. Is the VM error related to me unmounting the parity disks or does it have something to do with the other disk related errors?

 

 

I've also attached a more fresh diagnostics.

 

 

Thanks for the help.

closet-diagnostics-20230902-0834.zip

Edited September 2, 20232 yr by What's_a_Computer?

4 hours ago, What's_a_Computer? said:

Does the error long 'full of disk related errors' indicate inevitable system failure if nothing is done?

Not necessarily, but cannot see the start of the problem.

 

4 hours ago, What's_a_Computer? said:

Is the VM error related to me unmounting the parity disks or does it have something to do with the other disk related errors?

No, likely related to one of your cache devices having dropped offline in the past:

Sep  2 08:10:10 Closet kernel: BTRFS info (device sdd1): bdev /dev/sdd1 errs: wr 87055308, rd 359304, flush 1233992, corrupt 58988523, gen 0

Run a correcting scrub, also see here for more info and better pool monitoring.

I bought a H310 and moved the parity drives to that instead of the SATA ports on the motherboard pictured above.

 

If they stay synced for three weeks I'll mark this as solved. Thanks for your help @JorgeB! Your posts in other threads helped me a ton too.

The disks have been synced since I got the new H310 card. Looks like the issue was related to the SATA ports on the motherboard.