Identify disks?

cameronkeel

Dabbler
Joined
Aug 2, 2021
Messages
10
Got a question:

How do I tell which disk has faulted and which are degraded?

  • Disk 13983190838963152663 is FAULTED
  • Disk 18142404035722765029 is DEGRADED
  • Disk 13112144816528117927 is DEGRADED
  • Disk 9258244120143883618 is DEGRADED
  • Disk 13675505769364078550 is DEGRADED
 

jgreco

Resident Grinch
Joined
May 29, 2011
Messages
18,681
Got a question:

How do I tell which disk has faulted and which are degraded?

  • Disk 13983190838963152663 is FAULTED
  • Disk 18142404035722765029 is DEGRADED
  • Disk 13112144816528117927 is DEGRADED
  • Disk 9258244120143883618 is DEGRADED
  • Disk 13675505769364078550 is DEGRADED

How are your disks connected to the system? That's a lot of degraded disks. You start by looking for the cause. Did you cook the disks? Did you cook the controller? Did you drop it while running? Did you use a RAID controller?
 

cameronkeel

Dabbler
Joined
Aug 2, 2021
Messages
10
Not sure what you mean by cook. No, I haven't dropped anything while running.

I have 3 tower drive bays connected to a controller card. I know I have 3 drives that are pretty old Seagates, I'm thinking that is where the trouble lies, I just can't tell which drives this error is referring too.
 

jgreco

Resident Grinch
Joined
May 29, 2011
Messages
18,681
Cooking refers to the process of heating components, unintentionally, until they fail. This is a common cause of multiple components failing simultaneously, though others, such as PSU issues, bad cabling, etc., are possibilities too.

What is "a controller card"?

How do you expect us to help you decipher what's going on when we have no idea what your system looks like? We need more information, please.
 

cameronkeel

Dabbler
Joined
Aug 2, 2021
Messages
10

I have 3 Sans Digital TowerRAID TR5M+BNC - 5 Bay eSATA with a raid controller card. 14 ttl drives. Freenas running off ada14 (USB thumb drive).​

 

jgreco

Resident Grinch
Joined
May 29, 2011
Messages
18,681

I have 3 Sans Digital TowerRAID TR5M+BNC - 5 Bay eSATA with a raid controller card. 14 ttl drives. Freenas running off ada14 (USB thumb drive).​


Your TowerRAID TR5M+BNC is a port multiplier. Port multipliers are not expected to work. See


a raid controller card.

RAID controllers are extremely dangerous for use with ZFS. See


Basically your system isn't expected to work at all, though occasionally by pure luck things can work out for awhile. I would not trust this system with any valuable or important data that you wanted kept safe.

Your experience with the resilvering indicates that the system is not completely stable under stress, and ZFS likes to stress systems. You might have gotten away with it this time, but since the vast majority of resilvering involves reads, and you still managed to rack up some checksum errors (possibly as a controller hang, given that it seems to have considered a file damaged).

My best theory, understanding how port multipliers work, would be that one of your disks indeed faulted, had a read error that consumed excessive time and incurred a timeout/bus reset/something-like-that, and as a result the other four drives sharing that port also lost I/O transactions as part of the reset. I bet if you identify the four "degraded" drives, they're all in the same chassis as the "faulted" drive.

You need to get rid of the port multipliers and the RAID card.
 

cameronkeel

Dabbler
Joined
Aug 2, 2021
Messages
10
I think I mis-spoke, I think it is only a port replicator, no RAID card. I think/thought at one time this was in spec. I've been running this for 10+ years at my home, with no issues other than a few drive fails, but I knew that going in at the time, super cost restriction.

I would be interested in suggestions on a suitable system I could move all 14 disks too.

My real question though is, Is it possible to identify this drive:
  • Disk 18142404035722765029 is DEGRADED
 
Top