After about 48 total hours of data recovery with Data Rescue II, all 1.1TB of my data have been recovered. It is now spread across three drives, two 500GB G-raids that were both partially full and one 1TB Western Digital MyBook. Though I’d prefer not to have so many drives hooked up all over my desk, I am glad that I was able to save my data from oblivion.

What’s better (or worse) is that I’ve discovered the source of my problem. But I jump ahead of myself.

After recovering all my data and checking to make sure everything was ok, I went to do what any sane person would do in a situation like this—I ran DiskWarrior on the Drobo again. This time it finished! But, uh oh, it revealed the folder structure of the root of the drive, but only about 78GB of the 1.1TB I had on there was visible to the MacOS. Yet the Drobo and its utility Drobo Dashboard seemed to indicate that all the data were still on the drives even though DiskWarrior could not make it visible again.

So I figured that the shit was just too fucked up and since I’d saved everything, I formatted the Drobo. I assumed that formatting the Drobo would alleviate the problems I’d been experiencing since it was, obviously, a case of a severely corrupted directory structure, right? As an additional precaution I yanked the 500GB drive I’d salvaged from a failed G-Raid, leaving only the two Seagate drives I’d purchased specifically for the Drobo. Everything seemed ok after the format, the drive mounted, it had a name, all good. Then I tried to copy my photos onto the Drobo and after about 6GB, the copy slowed to a crawl and then hung, telling me that it would take an additional 140 hours to finish copying the remaining 109GB. What the hell? I thought. What is this, 1993? I stopped the copy since waiting 140 hours for 115GB it totally stupid.

I thought that maybe the Drobo needed a refresher, so I formatted it again. When formatting a Drobo, it gives you the option to chose a disk size of 1, 2, 4, 8, or 16 TB regardless of how much real physical storage you have. Perhaps some of the weirdness I’m having comes from the Drobo making the OS think that it is larger than it actually is? This time I formatted as a 1TB Drobo since I had two 1TB Seagates which yields 1TB of usable storage. Makes sense. The formatting tool indicated that the format with this particular drive configuration would yield 1 1TB partition. Perfect. But when the format was done, I had the 1 1TB partition and one brand new mystery 1 1TB partition that came from nowhere. I thought it must be a mistake so I formatted again and got the same results.

I went to the Drobo site to look for any help I could in the knowledgebase. For some reason I clicked a link to The Unofficial Apple Weblog to browse through as a distraction. For no real reason at all, just that I like that site and was interested in what they had to say about the Drobo. They were totally positive which was nice, but not helpful to me. I went to their main page and eventually encountered an article called Seagate continues to communicate bad news.


I figured it could not possibly refer to the server-class drives I’d purchased on recommendation from Lloyd’s Mac Performance Guide. I clicked the links to the two preceding articles.

The first was called, Tick tick tick… significant number of Seagate hard drives failing. It doesn’t go into too much detail. The real meat was in the second article.

Take two of the Seagate saga.

Now things get interesting. The article says this:

This is really important, because if it fails because you haven’t followed the instructions, your hard drive is offline. The data will be OK, but it will not be accessible. Seagate is saying they will retrieve data from problem drives, but that is a slow process and a gigantic hassle.

That sounds A LOT like what was happening to my drives, except I had the Drobo and Data Rescue to let me know that the data were still intact. The data WERE ok, but were not accessible. It was a GIGANTIC hassle and a slow process to recover. 2 for 2. I went to the Seagate page that has information about this shitshow and put in the serial numbers of the two drives I’d purchased. Guess what? One of them was in the batch of drive affected by this particular SNAFU. Fun! Let’s fix it!

Oh what is that? I can’t fix it on a PowerMac at work? I can’t fix it in a FW enclosure? I need to insert it into a MacPro or a Windows computer, neither of which I have or have access to? Oh, it’s because the firmware updater ONLY RUNS AS A TINY DOS EXECUTABLE ON INTEL MACHINES ON DRIVES THAT ARE ATTACHED VIA SATA!? Awesome! That makes it really easy to fix on my laptop computer with no accessible internal SATA ports. Super fun. And since I can’t fix it at work, and people have been getting refurbished drives back from Seagate instead of the new ones they’d sent, my only other real choice is to try and connect the drive via eSATA and a dock which I’ve read has worked.

Fingers crossed, people! We’ll get to the bottom of this yet!