Drives repetedly failing.. bad luck or issue outside of drives?

[bs.king]

Hello,

 

So I recently had my drives start failing out of nowhere. Initially drive 7 failed due to read errors, it had many pending sectors so I replaced it. When rebuilding the replacement, drive 1 stated it had errors. I paused the rebuild bit continued it at a later time without issues. Now the replaced drive 7 is disabled again and the SMART report seems fairly benign. Drive 1 is also giving errors again. Could this be a hardware issue? Could my drives be too hot (weather has been brutal the last 2 weeks)? Or is this just very bad luck with the drives?

 

Below are the diagnostics after the initial fail and after the subsequent one. Let me know if any more info is needed.

 

Thanks in advance!

mordor-diagnostics-20220822-1413.zip mordor-diagnostics-20220818-2337.zip

[JorgeB]

In the oldest diags disk1 issues look more like a power/connection problem, disk ended up dropping offline.

 

In the newest diags disk1 initially looks like a power/connection problem, but then it's logged as a disk problem, and the disk has pending sectors, so possibly problems with the connection and the disk, it might even be power problems causing the pending sectors, disk7 looks like a power/connection problem.

 

 

[bs.king]

6 minutes ago, JorgeB said:

In the oldest diags disk1 issues look more like a power/connection problem, disk ended up dropping offline.

 

In the newest diags disk1 initially looks like a power/connection problem, but then it's logged as a disk problem, and the disk has pending sectors, so possibly problems with the connection and the disk, it might even be power problems causing the pending sectors, disk7 looks like a power/connection problem.

 

 

Is there any way to know whether it is a power vs connection issue? It would make sense as these issues keep happening during rebuilds. I am running 8 drives through one PCI card, which seems like a lot....

[JorgeB]

Difficult to say for sure, check all cables, also make sure you're not splitting any SATA power to more than two.

[bs.king]

On 8/22/2022 at 3:06 PM, JorgeB said:

Difficult to say for sure, check all cables, also make sure you're not splitting any SATA power to more than two.

Hey so I believe I got the drives up and working, problem is that drive 7 was disabled for a long while and contents "emulated", when I rebuilt it, it was rebuilt with 1300GB instead of the 1750 GB that was on there before. Luckily I was able to pull the information off the old drive by sticking it in USB3 reader. While trying the rebuild, drive 1 failed and I haven't been able to get to the information at all - just read errors every time I try...

 

Is there any way to restore it by forcing the array to use either the rebuilt drive 7 with less data than expected, or by putting all the old data back onto drive 7? I am assuming no, since the drive was emulated for a while and probably had a small amount written to it. If not, is there any way to go about recovering data from the drive? I can see all the files, just not copy anything. Smart report is below with an older report

ST2000LM003_HN-M201RAD_S34RJ9CG306662-20220824-0128.txt ST2000LM003_HN-M201RAD_S34RJ9CG306662-20220823-1731.txt

[trurl]

46 minutes ago, bs.king said:

problem is that drive 7 was disabled for a long while and contents "emulated", when I rebuilt it, it was rebuilt with 1300GB instead of the 1750 GB that was on there before

Emulation allows any writes to the missing disk to be recovered, so you must have moved or deleted some files while the disk was emulated.

 

Start the array then post new diagnostics

[bs.king]

1 hour ago, trurl said:

Emulation allows any writes to the missing disk to be recovered, so you must have moved or deleted some files while the disk was emulated.

 

Start the array then post new diagnostics

Okay here they are.

ST2000NM0033-9ZM175_Z1X30WY1 (sdf) is a new drive I just pre-cleared, ST2000LM003_HN-M201RAD_S34RJ9CG306662 (sdo) is the failed disc 1 from previously.
I have the wrongly-rebuilt drive 7 and another drive with the original contents of drive not connected at the moment. Only really goal right now is to save whatever I can from the failed drive 1.

 

I really appreciate the help.

mordor-diagnostics-20220826-0050.zip

[trurl]

Don't see how you can have 2 missing disks with only single parity.

 

Post a screenshot of Array Devices, be sure to include all disks in the array and the full width.