Who the hell would pay $500 for a Drobo ?

What the hell is a Drobo, you ask ? Drobo is a little box that connects to your computer via USB, and houses from one to four hard drives. It does RAID-like things to protect your files from disk failure, and presents itself as a very simple, easy-to-use gadget.

And it costs $500, that’s not including hard drives!

Now I’m extremely biased because

  • 1. I’m a super geek
  • 2. I’m a storage freak and
  • 3. I’m a cynical bastard

But even after considering the non-technical users I still can’t justify the price tag on this thing. All it does is operate a proprietary imitation of RAID-5 on a cheap embedded processor and board, with a cute little web interface and a USB cable. It’s nothing revolutionary, and it’s something that anyone with even basic technical knowledge can build out of an old PC and free software like Openfiler, FreeNAS or the commercial (but inexpensive) NASLite. Sure, a FreeNAS system won’t plug over USB, and won’t look as cute on your desk, but who the hell cares ? Dig out an old Pentium-2, find a bunch of cheap hot-swap drive trays on eBay and knock your thrifty self out!

And spend the remaining $460 on hard drives, hookers and booze!

Both comments and pings are currently closed.

4 Responses to “Who the hell would pay $500 for a Drobo ?”

  1. Marti says:

    I think you are missing the point of the Drobo: it can work with any mix-n-match of different hard-drives. And it can be expanded by just adding a new hard drive (any size) or replacing one with a bigger one. On the fly. Classic RAID 5 isn’t that flexible and easy (ie: if you have a 3 x 500 GB RAID array, you can’t just plug an extra 250 GB in and hope to have 250 GB more capacity)

  2. Billco says:

    Oh, but I do get the point. They did use something different from classic RAID-5, allowing them to dynamically size and shuffle things around, but it seems like a half-assed solution to a problem that doesn’t necessarily exist.

    Mixing and matching also results in wasted space; if you have a 500gb drive and a 250gb drive, your usable space will be 250gb only, reducing it to a glorified RAID-1 mirror. Sure, it protects your data which is a good design philosophy, but if you’re going to drop this kind of money a storage controller, you’re probably not going to stuff it with random sized drives, which makes this feature practically useless.

    If the Drobo were priced at $150 instead of $500, my opinion would be different. At its current price point, it’s just a non-starter. We now have terabyte-sized drives on the market, and 500gb is presently the sweet spot. You might as well just buy two or more high-capacity drives, use the operating system’s built-in mirroring capability and skip the Drobo altogether.

    When your Drobo dies (and it will), you will lose everything, and the only way to get your data back is to buy another Drobo, because no other hardware or software is compatible with this custom solution. Just ask any network administrator about the last time they fried a RAID controller.

    In contrast, if you just use Windows’ software RAID, and your motherboard dies, you can plug your drives into any other PC and keep on truckin’ as if nothing had happened.

  3. Marti says:

    I’d agree that it’s a bit too pricy, however the ability and convenience of mixing and matching drives is extremely valuable. If you are running out of space, you just need to go out, buy the biggest drive you can afford and slide-it in. No configuration required at all, no reboot or downtime, no worrying of finding the same drive as the previous ones. Anybody can do it safely in 5 seconds, whereas growing a RAID 5 array with command line utilities isn’t for the faint of heart nor the average Windows/Mac user.

    Also technically speaking Drobo should not be compared to RAID 5 but rather to the new ZFS filesystem.

  4. Billco says:

    Well the thing is, ZFS isn’t even new. It’s been available in Solaris for about 2 years now. It does have some limitations that were apparently solved or at least kludged in the Drobo, but the whole reason why I’m such an asshole to this product and company is that I feel they’re targeting the wrong market.

    As a home user, I’d much rather blow $500 on a bunch of big hard drives and use software mirroring. This device is limited to 4 drives, which is puny since they could probably scale it to 8 or 16 drives at a very marginal increase in cost.

    A 16-drive Drobo, I could see myself paying $500-700 for such a thing… at that price point it would be directly competing with workstation-class RAID controllers and SAS solutions, and the cost of the Drobo would be amortized somewhat by the large number of drives. It would become an investment, rather than some half-brained gadget from yet another bunch of crackheads in California, a state which has firmly established itself as the MTV of internet technology.