Notifications
Clear all

Certifying 4 x 16TB disks in parallel on Thunderbolt 2

(@wisehaus)
Active Member Customer

Hi folks,

I have a question concerning the disk certification process in SoftRAID (v 5.8.3). I have 4 new 16TB disk housed in a ThunderBay 8 enclosure. I am in the process of certifying the disks before creating a RAID5 volume with them.

Since the certification process is going to run several days, I tried to certify all four disks simultaneously. However, I only seem to be able to run the process on 3 drives simultaneously. If I try to also add the fourth drive, I get read errors for that drive. The other three disks continue along happily.

I know the Thunderbolt 3 interface should be able to handle the data I/O for all fours disks, however, the enclosure is actually connected to an older Mac Mini which has a Thunderbolt 2 port so that might present a bottleneck.

So my question, should I be suspicious of the 4th disk itself or is it likely the reduced I/O bandwidth of the Thunderbolt 2 interface that could be causing the 4th parallel certification run to fail?

Thanks in advance for any advice.

Quote
Topic starter Posted : 25/05/2020 12:11 am
(@softraid-support)
Member Admin

There is no bottleneck, even with Thunderbolt 1.

Drives should never fail a certify, even if you did 8 at once, which would flood the bus.

This points to a hardware problem in your system. the trick is figuring out where.

It will take some research on your end, whether cable, drive, enclosure something. But I would want to get to the bottom of this, as it tells you that your system cannot reliably read/write data. A 4 disk Certify should never ever fail!

Is it any 4 drives, or just the last drive in the enclosure?

Did you try changing the cable or connections?

How long before you get the error?

ReplyQuote
Posted : 27/05/2020 10:09 am
(@wisehaus)
Active Member Customer

There is no bottleneck, even with Thunderbolt 1.

Drives should never fail a certify, even if you did 8 at once, which would flood the bus.

This points to a hardware problem in your system. the trick is figuring out where.

It will take some research on your end, whether cable, drive, enclosure something. But I would want to get to the bottom of this, as it tells you that your system cannot reliably read/write data. A 4 disk Certify should never ever fail!

Is it any 4 drives, or just the last drive in the enclosure?

Did you try changing the cable or connections?

How long before you get the error?

Ok, I have now done a bit of investigating based on your feedback.

First, I tried moving the disk which was giving the errors during the certification process to another empty bay in the enclosure. This seemed to be successful and certification process proceeded. At this point, all 4 disks are happily trundling along through the certification process in parallel. In fact, as I write this response, two of the disks have successfully completed and the remaining two are still running.

With the disks seemingly not to blame, I suspected the enclosure of course. So using one of the apparently healthy drives from the enclosure, I swapped this drive into the drive bay (Bay A) that the previously finicky drive occupied. The disk promptly failed to mount.

Moving this same disk to other unoccupied drive bays resulted in the disk mounting immediately and being recognized by SoftRAID. So only Bay A in the enclosure seems to cause issues.

At this point, the obvious suspect is that the enclosure itself is faulty. It is brand-new and still under all warranties. I live in Europe, however, so shipping it back for replacement will be a bit of a pain. Before I do so, is there anything else I should or could try?

ReplyQuote
Topic starter Posted : 28/05/2020 12:17 pm
(@softraid-support)
Member Admin

If this is an OWC enclosure, we do have a depot in Brussels for replacements. We have online chat, they can point you in the right direction.

But your test seems to have isolated the issue to the slot.

ReplyQuote
Posted : 28/05/2020 2:02 pm
Share:
close
open