6.7.2

 

I have no idea on how to proceed. Been using unRAID for approx 1 month. If I had to guess there is a hardware issue, maybe memory or motherboard. Is my data just fucked or is there a path to recovery?

 

Description: Yesterday received email notification that I had a Disk 4 that had errors and was offline. When I got home I checked and the drive was disabled. Had over 700 errrors.

 

Took the drive out and put in a new drive and began to rebuild. Said it would take approx 10hours.

Went back and checked approx 5 hours later and my server was offline.

 

Connected a Monitor and keyboard and no image on screen.

 

Performed a hard shutdown and powered back on, this time it was stuck on the BIOS. Shutdown again and unplugged USB to UPS. Powered up again and booted to unRAID successfully.

 

But immediately started to receive literally millions of errors.

 

Parity, cache, Disk 2, Disk 3 and Disk 4 are connected to SATA on MB. Disk 1, Disk 5 and Disk 6 are connected to an LSI SAS 9207-8i.

 

Additional Hardware:
Processor: 4790k
Motherboard: Gigabyte GA-Z97X
Memory: Corsair 2x8GB 1866 DDR3
PSU: Antec 850W
GPU: EVGA GeForce GTX 1050 TI

If you have a prompt login and type diagostics just to capture a log.

 

First step shut down and reseat all cables. The drive dropping off may have had a similar problem. You mention an lsi card is it seated properly in the mini sas connection? May be worth finding out if your motherboard has Marvell controllers. Having this much weirdness at once makes me think whatever the data connections are being driven from is having a problem with unraid. Rerun the data cables/swap. I'll leave it for the more experienced to give more ideas. Also to be sure you aren't having any brownouts correct? Hopefully using a UPS of some sort.

Edited September 1, 20196 yr by phbigred
Autocorrect issue

18 minutes ago, phbigred said:

If you have a prompt login and type diagostics just to capture a log.

 

First step shut down and reseat all cables. The drive dropping off may have had a similar problem. You mention an lsi card is it seated properly in the mini sad connection? May be worth finding out if your motherboard has Marvell controllers. Having this much weirdness at once makes me think whatever the data connections are being driven from is having a problem with unraid. Rerun the data cables/swap. I'll leave it for the more experienced to give more ideas. Also to be sure you aren't having any brownouts correct? Hopefully using a UPS of some sort.

thanks for the suggestions.  here is the diagnostics log. 

 

I have left it running and is about 36% complete on recreating the Disk 4 HDD.  Should i let it continue or just shut it down?    Not sure if I am causing more harm to keep it on or more shutting it down while it is trying to recreate disk 4.

 

edit:  and it has been on a UPS with other devices plugged in.  I haven't seen any issues with the other devices. 

 

I have only had the LSI for right at a week.  Didn't have any issues installing it or getting drives to show on it. 

tower-diagnostics-20190831-1925.zip

Edited September 1, 20196 yr by BobbyPoor_Boy

Disks 1,5,6 are all disconnected. Are these on the motherboard or the LSI? No point in continuing with a rebuild when other disks can't be read. All disks must be reliably read to reliably reconstruct a disk.

 

Looks like the webUI is working from your screenshot, so it shouldn't be necessary to get diagnostics at the command line as suggested. Just go to Tools - Diagnostics in the webUI.

 

Check all connections, power and SATA, then post a new diagnostic.

17 minutes ago, trurl said:

Disks 1,5,6 are all disconnected. Are these on the motherboard or the LSI? No point in continuing with a rebuild when other disks can't be read. All disks must be reliably read to reliably reconstruct a disk.

 

Looks like the webUI is working from your screenshot, so it shouldn't be necessary to get diagnostics at the command line as suggested. Just go to Tools - Diagnostics in the webUI.

 

Check all connections, power and SATA, then post a new diagnostic.

 

That would be correct.  1, 5, and 6 are all on the LSI.  All the other drives are connected to the on-board sata.  Attached is the diagnostics. 

 

Thanks for your help.

tower-diagnostics-20190901-0151.zip

Update: 

 

TLDR: the PCIE x8 slot on my motherboard is bad.  Moved controller to PCIE x4 and is booting without issue now.  All drives are showing up.

 

Further Explanation:

 

So I'm still not sure what happened with Disk 4, which is connected directly to the motherboard.  But user trurl had good advice and after unplugging/plugging in the various drives, cards and accessories, it looks like the LSI controller is causing the boot issues and disconnected drives.

 

about 50% of the time that the controller is inserted the computer won't POST.  This is with no drives connected to Controller. 

Sometimes it will post and start to boot to OS, but hangs on "starting SAMBA" and listing drives.  And sometimes if fully boots to the OS.

 

With the LSI Controller removed it boots to the OS 100% of the time. 

 

Moved the controller to the PCIE x4 slot and the PCIE x 16 (GPU uses this slot) and booted numerous times in each slot without issue.  So it looks like the PCIE x8 slot on my MB is bad.  Now booted and rebuilding the Disk 4.

 

Thank you everyone for your help.