I've now moved disk 8 onto a new controller and checked wiring, it's down rebuilding. Fingers crossed!

Disk 8 has rebuilt OK, but it reports Data-Rebuild finished with 244 errors. However parity is valid

Do you recommend any steps before I replace failed disk 7 with a new disk

I have suspicions that I started to have problems with disks 7 and 8 following a power cut, although I have a UPS

I've already replace one controller card with a spare that I had laying around, but I have now ordered 3 new LIS9207 controller cards to be sure as they are cheap to buy.

tower-diagnostics-20260512-1444.zip

23 minutes ago, clevoir said:

244 errors

Exactly this many lines in syslog with read errors for disk0 (parity). Since you have parity2 rebuild didn't need parity.

SMART attributes for parity look OK, but syslog said critical medium error.

Do Extended SMART self-test on parity.

Extended SMART test for Parity completed with no errors

I have now replaced Disk 7 and rebuilding the array

tower-smart-20260513-1539.zip

I now have disks 7 & 8 rebuilt, and will test the failed disks

I'll replace the control cards as precaution too

Thanks for your help

These are smart reports for the 2 failed disks, I am doing preclears on both of them now

I note for one of the disks it has 184 reallocated sectors, I guess this one is best thrown away?

WDC_WD60EFPX-68C5ZN0_WD-WX42D8445KLT-20260514-1606.txt WDC_WD40EFAX-68JH4N1_WD-WX22DB05D97H-20260514-1605.txt

1 hour ago, clevoir said:

one of the disks it has 184 reallocated sectors

I don't see that in either of those reports

Perhaps I have misunderstood the report but for the 6TB drive I see?

But also

5 minutes ago, clevoir said:

I just look at the RAW_VALUE, which is 0.

I can see that that the Reallocated Sector count is 0, but I'm concerned that the Number of Realloc. Candidate Logical Sectors is 184?

I have never seen this heading before, and Google doesn't give a lot of information on it either?

44 minutes ago, clevoir said:

Candidate Logical Sectors is 184?

What do you mean by this? Both SMART reports look perfect to me.

Hi, I am probably overthinking as both drives show 0 reallocated sectors🙂

I have never seen this heading before, is this 184 sectors that can be reallocated if needed, as the current relocated Logical sectors is 0?

I've tried searching on Google but can't find anything about Candidate Logical Sectors

Edited May 15May 15 by clevoir

You can ignore that.

Thank you

After an uneventful few days I have 2 disks failed again!

Of the disks that failed last time, these were replaced with larger disks, and I carried out 2 preclears and 3 extended SMART tests each of the failed disks and found no fault. The parity disk 2 that has now failed, it previously failed and I carried out a extended SMART test and found no fault. So I have strong suspicions that it's not a disk problem!

My problems started after a power cut, all connections data / power were OK. I wonder if a have controller problem, I read that DELL PERC H200 cards don't like voltage fluctuations?

I've started the array without the failed parity disks, added them back in and the array is rebuilding.

After last time I have bought new HBA cards and new SFF-8087 to SATA wiring too, I was going to swap this out at the weekend.

My plan now is to get the array rebuilt OK and then immediately swap out the cards & wiring

See attached diagnostics

tower-diagnostics-20260519-1414.zip

Disks just dropped offline, looks like a power/connection problem.

The failed disks were shown as unassigned devices, but when i tried to reassign them to the array, the disks were shown but as soon as i selected them their names disappeared as if they did exist.

I saw this before, but this time to rebooted the server, upon reboot i was able to reassign the disks OK

So i can see them dropping, i guess either because it's power or the HBA is dropping them

I will strip down my server and check all connection when i replace the controllers / cabling