Hard Drive Errors

[Ellimist]

Hi All,

Just looking for some advice here. I have a few drives with errors on them. Freenas hasn't removed them from the pool yet. I'm wondering should I fix them or should I replace them?

It would seem that the unreadable sectors I just can't seem to fix. I followed the instructions to do so that were on the forum but for whatever reason it just doesn't seem to work.

The error counts I'm not sure about.

The drives are all in different Z2 pools so its not an issue of losing the volume at the moment.

Would a wipe of the affected drive resolve any issues? or should I just go with drive replacement?

.

Thanks for the advice.

Cheers,

 

[nojohnny101]

Yikes you have 3 separate drives throwing up errors? I hope you have a backup of your important data.

First you should look at the smart information for each mentioned drive.

Next, look for specific parameters that indicate a drive failure.
Here is a great resources written specifically for hard drive troubleshooting like you are experiencing:
https://forums.freenas.org/index.php?resources/hard-drive-troubleshooting-guide.17/

 

[Ellimist]

Thanks thats a brilliant resource.

Yes I have a complete backup of all the data on separate hard drives. Hence why I'm not so stressed about the drive issues. hehe.

 

[rs225]

Any of those drives would be expected to fail a long test, until the bad sectors are overwritten/relocated and the problem 'goes away.' One way to force it is a full over-write of the entire disk.

If any of them are failing short tests, then definitely replace. If the count of problem sectors is rising, then that would also become a replace.

If the pool passes a scrub with no problems and the error counts aren't rising, then at the moment, there is not an urgent problem. However, paranoid people would not have 3 drives with counts in one pool. ;)

 

[joeschmuck]

@Ellimist
The best advice I could provide you it to take quicker action on noted failures. Run the SMART Long tests per the guide you were linked to and if they fail, take appropriate action. @rs225 mentioned writing data to the entire drive (or more so the the failed LBAs) to force the failing sectors to be locked out from use however odds are the problems will return as more media degredation occurs. If the drives can be RMA'd then I'd do that before performing that action. If the drives are out of RMA then you could try it but I'd personnaly replace them with good drives. Well that is what I'd do.

 

[Ellimist]

Yeah I got the drives yesterday to replace them. I have scrubs and smart checks set up per cyberjocks guide.

https://forums.freenas.org/index.php?threads/scrub-and-smart-testing-schedules.20108/

 

[joeschmuck]

I prefer to run a daily short test and a weekly long test.