Impossible Read/Writes on Parity Drive

[n8girard]

I have been having hard disk failures multiple days this week resulting in full drive rebuilds on several BRAND NEW 10TB IRONWOLF SEAGATE NAS drives. I know it's not the drives, though they have replaced the first ones already under warranty just to be safe. It's following a pattern though and this has happened at least 4 times already with the same symptoms:

1. One of my drives goes down and gets my attention.
2. I have a Parity drive that reads 22,533,332,543,025 Reads and 22,533,332,543,360 Writes (which is impossible because since booting it's only been 3hrs.) Note this has also happened within just a few minutes of me checking the server which is also impossible. See attached screenshots:
3. Inevitably there are some drives with errors on them and it stops those drives automatically.

Troubleshooting steps taken:

1. Completely removed drives with error on them and replaced them with brand new drives. These new drives fail within a week with same symptoms as before.
2. Replaced system Cache drive with a new one. No errors on it to date but problem still occurring.
3. Replaced the Flash drive with a new one and moved my registration key over. 
4. Upgraded my OS and 3 drives failed to be recognized: Parity, Disk 3 and Disk 4, all three of my 10 TB Ironwolf Seagate drives.
5. Restored Flash drive to previous state and old OS v 6.4.1 and now back to only one disk failed, others were recognized. Rebuilding Disk 4 now.

 

I also included my system info. 

I believe my next step is to replace the Parity drive and rebuild it. Any other recommendations? Anyone seen this behavior before?

[trurl]

Go to Tools - Diagnostics and attach the complete Diagnostics ZIP file to your NEXT post in this thread.

[n8girard]

Attached is the diagnostics file.

plex-diagnostics-20201121-0105.zip

[JorgeB]

I have a Parity drive that reads 22,533,332,543,025 Reads and 22,533,332,543,360 Writes

This is a symptom of a drive dropping offline, diags are after rebooting but you're using a Marvell controller and those are known to cause that and other problems, all 3 disks showing errors on the GUI are unsurprisingly connected to it.

 

 

[n8girard]

@JorgeB This is likely due to a PCIe SATA card I installed to expand my SATA capabilities. Evidently it's garbage. Any recommendations on a good quality card that won't drop my drives?

[JorgeB]

How many ports do you need? There aren't many good 4 port options, there's the 5 port JMB385 controller, though most times you can get an used 8port LSI HBA for about the same.