New Hard Drives, New Controllers, Many Read Errors

[BoySherman]

I think I've tried everything I'm stumbled across in the support forums, but I'm stumped.  

Back story; I had an old broadcom SATA card.  I was experiencing VERY slow transfer on parity sync (< 1mb/s).  Funny stuff started to happen with docker config files, disappearing files etc etc.  So, to remedy this, I purchased 3x "serial Attached SCSI controller: Broadcom / LSI SAS2008 PCI-Express Fusion-MPT SAS-2 [Falcon] (rev 03)"

After installing these cards, parity sync is great, and back to what I think is expected (~150mb/s).  However, *some* of my somewhat newer/larger hard drives are throwing up errors at an alarming rate.  Some of them even after a complete format (there is nothing of real importance, so I was happy to just delete data in some cases).

I have done the following (and probably more);

* ran a checkdisk -L on every disk.

* completely formatted some (still seeing errors)
* Ran a memtest (clean).
* removed memory modules down to a single chip for now
* Reset all of the controllers in their slots, ensured sata cables were all secure.
* If a hard drive failed, moved it onto another controller where I know the connection is sound (I have hot-swappable drives)

* seems like hardware faults, but I don't understand how it could have occurred to many at once, the odds just don't add up.  (especially after formatting).  

Diagnostics attached.  Any ideas?  
 

tower-diagnostics-20220213-1026.zip

[BoySherman]

I should make note of one thing I forgot... all of the drives power are daisy changed off of a single supply cord (with splitters).  Not sure if that's of any concern.

[JorgeB]

1 hour ago, BoySherman said:

Not sure if that's of any concern.

It is, see here for a recent discussion about that.

[BoySherman]

Boom!  Yup, that was it.  Split the 6x4 backplane into two, and the funny behaviour ceased... thanks!