Hello All,  

I swapped an old 4TB WD Blue for a shucked WD easystore 12TB.  Normally, it takes between 2-3 days to complete an array rebuilt or parity check.  The estimated time for rebuild is bouncing between 3 and 16 days, mostly staying at towards the 16 day side of the spectrum at about 15MB/s rebuild speed.  

I'm not overly concerned, more curious if there is anything I should dig into, or if it's might be something harmless like a large number of small files slowing the rebuild.  Dual E5-2660 v2 @ 2.20GHz [CPU utilization averaging 8%] with 64GB ECC RAM [25% Used].

I've gone attached a fresh diagnostics download.

Thank You 

 

PS.  22 disk+2 disk parity array, 201TB array, with mixed 12TB and smaller disks. Supermicro X9DRi-LN4+/X9DR3-LN4+ , Version REV:1.20A

   

seine-diagnostics-20221016-1831.zip

Edited October 17, 20223 yr by Grrrreg
added information about array and disk sizes

Looks like rebuild had only been going for a few minutes when diagnostics were taken. Let it run a while and get new diagnostics if speed doesn't improve.

 

How have you managed to use 87G of 150G docker.img? 20G img is often more than enough. And never seen anyone need more than default 1G libvirt.img.

 

The usual reason for filling docker.img is an application writing to a path that isn't mapped.

 

 

There's something reading from multiple disks during the rebuild.

Thank you trurl and JorgeB for reviewing.   It did finally stabilize at normal speeds after a while, now at 13 hours remaining.  I've just never seen it take so long before.   

 

@trurlThank for pointing out the docker.img size and libvirt.img.   I had about 35GB of orphaned images I hadn't paid attention to, deleting those brought it down to the low 50GB's.  I need to do some additional clean up of other duplicate/test containers. Thank you

@JorgeBI did stop several dockers thinking they might be adding to it.  Next time I'll stop all but critical containers.  Thank you