News WD Fesses Up: Some Red HDDs Use Slow SMR Tech

[King_V]

On the plus side, it looks like Seagate has clarified, at least for their NAS drives:

Seagate says Network Attached Storage and SMR don’t mix

Seagate’s IronWolf and IronWolf Pro NAS drives are all certified SMR-free.

arstechnica.com

 

[clintercell]

Sorry to hear that.

Please let us know how it turns out, especially whether SMR was responsible for the problem.

Good luck.

Thanks! Yesterday I was able to get help from Drobo's email tech support. I've heard horror stories about working with their support team, but I'm happy to say my experience was a great one.

Eric from Drobo was very responsive and invested in helping me get my data. We probably exchanged 6 emails yesterday working through the issue. It's not fully resolved yet, but we're still working on it.

Current status: We were able to get the drives back online by launching the Drobo in READ ONLY mode. I was able to then pull my critical data off of the Drobo onto other drives. I consider myself lucky and it is another reminder to always back up important data immediately.

Here is what we've noticed through troubleshooting the issue:

When the drobo boots without the 6TB RED Drive, it is consistent. The PC detects it, but it is inaccessible as the full raid isn't available AND one of the 3TB drives is marked as failed. It seems that the Drobo successfully rebuilt its raid for the 6TB hard drive.

When the drobo boots with the 6TB SMR hard drive, it is erratic. It can boot with different warnings and status lights - but it either connects to the PC as critical and inaccessible - or it tries to connect to the PC and malfunctions (Windows gives a USB device error). It can even cause the PC to bluescreen.

Here is what seems to be happening:

The Drobo rebuilt itself for the 6TB drive. After rebuilding itself for the new drive, it flagged one of the 3tb drives as "failed" (still not sure on the cause of this). Because of the "failed" status of the 3tb drive, the Drobo needs all the other drives available to get the data.

When in read/write mode, the SMR drive has unreliable or strange speeds, so it gets kicked from the Drobo and the drobo marks its bay slot as "empty." When in READ ONLY mode, the speeds are more reliable, so the drive doesn't go invisible or get booted from the Drobo.

I'll post another update once the Drobo diagnostics team comes back with more info. Right now the problem is clearly the 6TB drive. Now the question is, it is a bad drive on its own, or are the speeds of the SMR read/write causing the issue.

 

[bit_user]

Thanks for the update.

The Drobo rebuilt itself for the 6TB drive. After rebuilding itself for the new drive, it flagged one of the 3tb drives as "failed" (still not sure on the cause of this).

It could be due to an unreadable block on that drive.

Does the Drobo give you any way of querying the SMART logs of the "failed" drive? Otherwise, when it's all said and done, you can pop it in a PC and query it.

 

[geogan]

Funny enough I just put a 10TB Seagate in the machine to replace a full 6TB WD Red drive... so this 6TB WD Red drive was sitting on the desk beside me when I read this article 😆

Is it all WD Red 6TB drives that use this? How do we find out?

I still have another one in the machine.

Drive I took out is WD60EFRX (May 2018)

Drive in machine is WD60PURX - actually one in machine is getting bad - only 96% health (SMART Current Pending Sector Count reported warning status). It has 748 days power on time.

I meant to replace this but did wrong one by mistake 😵