Drobo V2 looks to have lost data while all green

Woke up this morning and my Drobo volume wasn’t mounted. Reboot everything, still won’t mount. Opened a ticket with an attached diagnostics file. Noticed there was a new firmware level (1.3.6), did the upgrade, it applied, still won’t mount.

Ran disk utility which sees the drive. Dashboard sees it as all green also. Can’t force mount the disk in disk utility. Tried verify disk. Lots of errors. Tried repair disk, disk is unrepairable.

This is pretty much the last straw for me, it’s going in the bin. I reckon what happened is the previous ticket showed bay 3 and bay 4 drives (WD20EADS) having spurious problems. Drobo was doing random short rebuilds to handle it but it never said what was happening, i.e. two drives were giving problems or having timeouts. Everything stayed green.

At Drobo supports request, I replaced the drive in bay 3 on Friday and that went well. I’m waiting for the advance RMA WD drive to arrive this week to replace the bay 4 drive figuring with 3 good drives, I’m safe even if bay 4 goes but obviously, that was optimistic.

How can a RAID device lose data without a reported media failure? Obviously, not very happy right now. Dashboard was running overnight and there are no emails suggesting a media failure. It’s hard to spin this in a good way. It’s attached to an iMac using USB, the cable is fine (firmware worked, everything else works, it’s attached directly to the iMac, no hub). This is a vanilla setup.

Case#: 100301-000006

doesnt sound promising, but best of luck & keep us informed!

Repair disk failed in disk utility, diskwarrior says the volume is toast so I guess thats it, I don’t see what the drobo support guys can do at this point besides provide feedback to development that the device can lose data without reporting a failure…

i guess its perfectly possible something else could have caused it, some corruption somewhere along the bus. evidently drobo thinks it is still redundantly protecting whatever corrupted volume information is on there.

i woudlnt give up too quickly (although it does look bad) wait and see what support suggest when you send in your logs.

or see if any other mac guys can offer more advice (im 100% PC)

I got nervous when I was getting the random disk rebuilds so I stopped using the Drobo as my main drive. It was too slow anyway. So, losing the data on it isn’t a big deal now. I reckon I’ll pull the bay 4 drive now and RMA it and then just use the Drobo as a time machine drive until I buy a QNAP or similar. Bay 1 and 2 showed no issues, bay 3 is a new drive as of Friday.[hr]

I guess so but I have other external drives connected to the same iMac with no problems. I may grow a third arm also given enough time. Razors occam…

If there’s a way to check for USB data communication errors, that’s where I’d start.
I’m PC too, so I don’t know a lot about recent Mac OS, but if there’s an automatic update function like Windows has, that could also be a trouble if some driver or system-level component got updated.

If Drobo says Green, then the integrity of the drives should be fine, but that means nothing for the integrity of the data. Just means there’s no “holes in the paper” - doesn’t stop something from scratching out the drawing.

Keep us up to date and I hope some resolution can come from things.

I think you’re grasping at straws there. I have other USB drives on the mac (even connected through the dreaded hubs) for even longer periods of time, they are my main drives also and no problems at all.

Support just got back to me but basically, all they are saying is yup it lost data. Try disk utility or diskwarrior and I’ve already done that so opening the ticket told me nothing I didn’t already know, the data is corrupt and it only seems to happen on my Drobo. Maybe MacOs has an “if(device==drobo) then do something crazy…”

Bad USB cable can cause transfer troubles. Another user mentioned that on another thread.
I’ve had it happen before with Firewire and USB, though not with Drobo.

If I’m reading your other threads correctly, you were encountering spontaneous relayouts prior to this and you sent your logs in. What did support say then?

DisKs inbay 3 and 4 had bad sectors. This is why I replaced bay 3 last week. Your point on USB is hard to believe also given USB includes crc checking in the protocol so the data sentorreceived is guaranteed correct. Not being able to senda packet would show up as an io error in macos which i’ve never seen. It’s looking more and more like a hole in drobo. Data was corrupted and not detected which us what’s making me rethink whether or not to trust the thing.

Call me cursed, but I’ve had a wide array of data issues. I once had two 3ware RAID controllers that worked perfectly fine on their own, but when put into the same system, one would develop random bit-shifts. Took a lot of copying of large files back-and-forth before I determined what the problem was. That’s part of the reason why I’m paranoid about any large-scale data copies and do full CRC-checks when I migrate data.

Anyway, back to your problem, I’m confused…

Sometime prior to last week, Drobo red-lighted the drive in Bay 3.
So last week you replaced the drive in Bay 3.

Relayout started and completed?
Any spontaneous relayouts after Bay 3 drive was replaced?

But now, both the replacement drive in Bay 3 and the drive in Bay 4 have failed, causing data loss?

Thanks, just trying to get a handle on the exact events that led to the current situation.

Nothing was redlighted. I had been getting dynamic rebuilds. A rebuild which lasts maybe ten minutes and then all green again. Bay 3 and 4 were the cause. Both were giving bad sectors occasionally causing the short rebuilds.

Drobo analyzed the diagnostic file and suggested pulling 3 first, then replace 4. I pulled 3 and replaced with a 1.5tb drive and it failed last night after it went all green on Saturday.

The replacement was a brand-new 1.5TB drive?

Yes. Could the gods have conspired against me and given me a duff drive? Possible but the rebuild worked after I put it in

It’s unfortunately possible… barring human interference (including “slow cooking”) they do have a tendency to die either early or late in their MTBF. I had laptop drive start screeching between day 4-7 of use.

i’d had a worrying number of drives that were useless straight out of the antistatic bag. my favourite was a 2TB hitachi which just screamed as soon as power was applied… it sounds like it was forceably dragging its heads over the platters… or someone had put sand in its bearings… it was NOT a pretty sound!

Yeah, that’s what the 2.5-inch I had did!! It was a horrible scraping/screeching sound.

im guessing it is probably a result of a knock during shipping