Errors on Multiple HDDs & Missing Disks

[Derrikdj]

I've setup a new Unraid server that I've been running for a few weeks now. I've been in the process of migrating my data & HDDs from a Windows machine to my Unraid array. I would transfer the files from an HDD to my Unraid array then preclear the empty drive before adding it to the array. I've been repeating this process until all of my drives will be added to the arrray. I've added 9 drives so far, but 4 drives have been disabled due to errors during data transfer. Two of the drives that have been disabled are brand new Toshiba MG08ACA16TE 16TB drives that were successfully precleared with pre & post reads. The other two are drives that had been running fine on my Windows machine for a while. It's hard for me to believe that I have 4 bad drives. The array drives are in a NetApp DS4246 shelf with a NetApp X2065A-R6 Controller that I bought both used. I'm concerned if maybe the shelf or controller are the root cause of the errors. I've attached diagnostics from the last two occurrence of errors. In one instance two drives had errors at the same time. One was a Parity drive.

 

My other issue I'm not sure if it's related or not, but after a shutdown or reboot different drives would be missing after booting back up. It seemed random because a different drive or two would be missing each time. At first I would keep rebooting until all the drives came up, but as I added more drives to my array at least 1 or 2 drives would be missing. Soon I discovered that if I removed a drive for about 30 seconds then reconnected it in it's slot, UnRAID would recognize the drive. Once again I wonder if this is an issue with the shelf or controller. 

 

Thanks for any help

unraidsvr-diagnostics-20200714-1614.zip unraidsvr-diagnostics-20200717-1511.zip

[JonathanM]

Power and supply wiring to the backplanes / drives?

[Derrikdj]

6 minutes ago, jonathanm said:

Power and supply wiring to the backplanes / drives?

The shelf has dual power supplies & are plugged into a UPS.

[JonathanM]

Just now, Derrikdj said:

The shelf has dual power supplies & are plugged into a UPS.

What about the DC path of the supplies to the drives? All tight connections and sufficient gauge and quantity of wires?

[Derrikdj]

9 minutes ago, jonathanm said:

What about the DC path of the supplies to the drives? All tight connections and sufficient gauge and quantity of wires?

As far as digging into the inner workings of the disk shelf is beyond my expertise or comfort zone. Do you see something in the diagnostic logs that would indicate the drives are disconnecting or losing power? I've never had an issue where a drive is missing while the array is running. It's only occurred right after boot up. Once a drive is recognized they are stable until a shutdown or reboot.

[JonathanM]

Sorry, I have no first hand experience with disk shelves, all my servers have the drives local to the motherboard / hba.

 

Maybe @johnnie.black has some insight, but it may be several hours before he checks back in.

[JorgeB]

NetApp controller uses a PMC Sierra chip, and IIRC I've seen issues with those and Unraid before, if possible get an LSI HBA instead, though you'll to get need a different cable.

[Derrikdj]

1 hour ago, johnnie.black said:

NetApp controller uses a PMC Sierra chip, and IIRC I've seen issues with those and Unraid before, if possible get an LSI HBA instead, though you'll to get need a different cable.

So you think the card is causing the errors?

 

Would you recommend I replace it with this card:

LSI 9200-8e 6Gbps 8-lane external SAS HBA P20 IT Mode

https://www.ebay.com/itm/LSI-9200-8e-6Gbps-8-lane-external-SAS-HBA-P20-IT-Mode-ZFS-FreeNAS-unRAID-NoROM/163534822734?epid=28034148027&hash=item26136f5d4e:g:5sEAAOSwdwlcX2E3

 

[JorgeB]

40 minutes ago, Derrikdj said:

So you think the card is causing the errors?

I suspect that it might be the problem, PMC controllers are very rarely used in general and with Unraid specifically, also the driver is likely not as mature and well tested as the LSI one with Linux.

 

42 minutes ago, Derrikdj said:

Would you recommend I replace it with this card:

That would be a good option.

[Derrikdj]

5 hours ago, johnnie.black said:

I suspect that it might be the problem, PMC controllers are very rarely used in general and with Unraid specifically, also the driver is likely not as mature and well tested as the LSI one with Linux.

 

That would be a good option.

Thanks. I've ordered the LSI card & cable for it. I'll Let you know if that resolves the issue once they arrive & I get them setup.