SOLVED Badblocks - 1 drive twice as slow as all others

[Stephen2]

Hi all, am up to a burn in phase, using the methods outlined:
https://forums.freenas.org/index.php?threads/how-to-hard-drive-burn-in-testing.21451/

There are 7 3TB drives attached via backplane SSF 8087 -> 4 SATA to the LSI controller on the motherboard (flashed to IT mode)...

After 5 1/2 hours, I've just checked the progress... da0 -> da5 are between 80%-85%.... da6 however, is at ~42%

Seems unusual... anything to worry about do you think?

Edit: should have mentioned, badblocks & all smart tests reporting no errors so far on this slow drive.

 

[Robert Trevellyan]

Seems unusual... anything to worry about do you think?

Could be. A drive can be so slow as to be useless without failing any other tests.

 

[joeschmuck]

After 5 1/2 hours, I've just checked the progress... da0 -> da5 are between 80%-85%.... da6 however, is at ~42%

How about being a little more specific... Are all the hard drives identical make/models? What test is being performed where you are doing this comparison, did you start the test of all the drives about the same time, and is this one drive always running slow for all the tests?

Lastly, you could start the test again but ensure you power off the drives first to ensure they are not doing some test in the background. Explanation: If you accidentally started a SMART Long Test on the suspect drive and then started running the Bad Blocks test, it could lead you into thinking the drive is slow. Cycling power will stop any testing in progress.

Lets say you repeat the tests and da6 is still slow... Then power off the machine and swap the SATA cables between da6 and any other drive to see if the problem is still there. If the problem is also with the SMART test, then it's not a SATA cable becasue SMART is all self contained with then drive.

 

[Stephen2]

Thanks everyone, these are good questions, and some good info and debugging techniques.

I didn't really know what info to provide, but looking back "All the same size/manufacturer" was an obvious one!! Yes, all 3TB WD Red NAS drives.

And "Turn it off and on" is a classic for a reason - it has worked again.

@joeschmuck - thank you for the solution ideas, you were right in that I look like I'd confused myself with TMUX or something and must have been running two things at once on the single drive.

After turn off/on, and starting again with badblocks on all at once, all drives are pretty much keeping up with each other.


Thanks heaps everyone.

STATUS: Solved.
CAUSE: User error, running 2 tests at once on same drive = dumb and makes those tests run slowly

 

[joeschmuck]

Glad I could help. I like to assume nothing because when I do, I can give bad advice. Of course even if I have all the information it doesn't prevent me from giving bad advice, sometimes I say stupid things.