I have a few Seagate IronWolf drives (12tb) in my array (mostlyr HGST 4tb drives), but the Seagate drives keep running into parity issues. This is even after swapping for another LSI HBA 9300-16i card updated to the latest firmware. The server is running a Threadripper 3960x with 128gb of ddr4 ram (no-ecc).

 

I would do an 18hour parity sync to rebuild, attempt to unbalance, and move files off the drive, but then ithe array would run into a parity error again on those Seagate drives. I see a lot of errors like this:

 

This seems to happen on other Seagate drives I've plugged in. Some of which were only a few months old. Smart results report nothing. I've also wiped it on my PC and ran surface tests on it. They all report nothing wrong with the drives.

 

Could there be something else at play here I am not seeing?

sarva-diagnostics-20241204-2249.zip

Edited December 30, 20241 yr by S0ulDrag0n
clarification

If the parity drives are larger than your biggest data drive then it is possible that when you added them the part beyond the size of the largest data drive did not get properly zeroed (due to an Unraid bug), 

 

If this is where the errors are being reported then you need to do one correcting check to get the situation corrected and then after that you should get 0 errors in the future.

 

If this is not the case then let us know.

There are errors with multiple disks, looks more like a power/connection issue.

11 hours ago, itimpi said:

If the parity drives are larger than your biggest data drive then it is possible that when you added them the part beyond the size of the largest data drive did not get properly zeroed (due to an Unraid bug), 

 

If this is where the errors are being reported then you need to do one correcting check to get the situation corrected and then after that you should get 0 errors in the future.

 

If this is not the case then let us know.

 

Both my parity drives are 12tb, not sure if they're off by a bit. I did run into an issue with one of my parity drives which I've taken out for now. The issue I am having is mainly with one disk (disk2).

It will error occasionally. I did notice something that I am not sure whether it is related. When it does error out, I can't rebuild immediately or it'll just run into an error again. I have to wipe it first. To make things go faster, I plugin my other Seagate 12tb drive to try and rebuild, while I wipe the one that errored on my desktop. I just write 0's to everything. The thing I noticed is after the error, the drive will typically not get detected immediately by my machine and tool. I have to unplug the device and plug it back in for the device to become recognized. This occurs on the drive I swap in when it errors out too. Both are the same brand and model.

 

10 hours ago, JorgeB said:

There are errors with multiple disks, looks more like a power/connection issue.

 

I think this was mentioned before. What I can try is directly hooking up the HDDs to my second power supply (which has more than enough capacity right now) and see if I run into the same issue. I will report back once parity is done rebuilding.

Edited December 5, 20241 yr by S0ulDrag0n
Removed irrelevant quote

Parity sync is done and it looks like it's stable. It seems like a power splitter I was using might be the cause. Not sure if it's just not capable of the power output or became faulty over time as this did not occur before when I had fewer low capacity drives. I do have a 450W secondary power supply powering all the drives now and will be switching other drives over slowly. Will keep you guys posted if something does happen.

Edited December 6, 20241 yr by S0ulDrag0n