Drobo

Disk failures - why are they happening so often

I have been using Drobo drives for years with WD Reds and have randomly suffered from the Dashboard reporting disks as Failed with the red light flashing. This only happens after I correctly eject the drive, power the Drobo 5D3, 5C or 5N down, and re-start it. Is this normal behavior, please?

To fix the “failed” disk I remove it, erase it and stuff it back in the same Drobo slot, so it thinks it’s a brand new working drive. Within 24 hours all disks are working properly again.

I make it my practice now to never turn my Drobos off as the above procedure is making me frustrated.

Does anybody have any helpful advice, please?

What age WD Reds? New Reds are problematic, WD changed how they record data internally, they use SMR which Drobos don’t like, see note 2 on this page all kinds of strange stuff happens. Including apparent failures during rebuilds, very bad news for your data if that occurs.

Make a note of the models listed here
Then go to Dashboard > Status
Pick Drive Information in the dropdown on the status page,
Select each drive in turn, if the first section of the numbers in the
Product ID field match the numbers on the linked page you’ll need to replace them with either
Red Plus, or some other drive which uses CMR rather than SMR. If it is that you could try raising an RMA with WD, though your mileage may vary.

If it’s not that, then I’m not sure what you should try short of raising a support ticket & sending diagnostics to Drobo themselves.

Thanks for your really informative reply. The WD reds are 4 to 6 years old so they are the reliable CMR type.

It might be that the drives really are EOL, however they have been going randomly “faulty” for several years, triggered by powering up the Drobo - but they have always been capable of being repaired by reformatting.

As the Reds are out of warranty, my only option would have been with Drobo customer service but from what I have been reading recently this would be a waste of time, especially as my 5C is out of warranty. Do you think it’s worth a try?

I really can’t say when it comes to support, depends how you view the price of a one-off incident vs what you think the chances of a resolution are.

As for the drives, when the next one gets kicked out, investigate it thoroughly, run Data LifeGuard long test, check out the SMART data, see if that’s showing reallocated sectors or other early warning signs, act accordingly. If it’s showing signs of wear/age/deterioration, you have the cause, if not then the Drobo may be getting flaky.