Hi,

 

So i had a defunct drive in my data set, i ordered a new one and precleared it with the preclear plugin.

This is a 3 TB disk and took about 17 hours to complete the preclear.

 

I followed the steps here  on how to replace a broken drive.

The rebuild started with around 200 MB/s and around 50% it went down to around 50 MB/s and even 30 MB/s

 

But i cant seem to figure out why, i dont have any errors on my parity disk.

 

The parity disk i have is an 8 TB WDC WD80EZAZ.

 

And all my data drives are TOSHIBA HDWD130.

 

Can someone put me in the right direction on why this rebuild is taking ages?

Is the parity drive to slow? But that would be strange because the first 50% is going at 200 MB/s.

 

I have also added the diagnostics.

 

Thanks.

 

Kind regards,

Dion

tower-diagnostics-20201230-2121.zip

Do you have anything else causing other activity on the disks besides the rebuild?

I had 2 containers active, 1 is zone minder and the other one is home assistant. Both where writing to the cache drive and not to the data drive.

 

There's something reading from disk2 in the diags posted.

some fade away is normal. but from 200 to 50 it is probably some other activity using the drive. 4TB rebuild took me 8 hours on the HGST disk drives for me starting at 200 MB/s ending at 100 MB/s corresponding with below disk speed test.

 

if you want to see what data is being written to your shares type

lsof | grep mnt

in the terminal console

That would be zoneminder i guess, next time i will disable that also (although its my security camera's) and see how it goes.