Should be arriving next week or the week after that… but continue reading for a more detailed discussion on the gadget.
It’s basically variation of a RAID 5 system implemented into a stylish box. But it takes away all the hassles that come with managing such systems - so it’s basically idiot-proof. Also, it has a bit of future-proofing thrown in for good measure, as described by the quoted text below:
There’s a theoretical 16 terabyte limit (when disks technology allows) but this limit is one of the file systems employed, and not the Drobo hardware itself.
The cool thing about Drobo is that unlike a RAID 5 setup, you aren’t hindered by the size of your smallest drive. The typical storage capacity of a RAID 5 array is 
where N is the number of drives you have and Smin is the storage capacity of your smallest drive. So if you got an 80GB + 500GB + 1TB setup, you’ll only have 160GB (80x2) available. With Drobo on the other hand, the same scenario above would give you 580GB instead of the typical 160GB of a RAID 5 setup.
The safe formula to assume when thinking about how the drive space spans seems to be the summation of the storage capacity of all drives that are less than or equal to (but not including) the biggest drive. Or like the image I made below:
So here are quick real-world applications of the formula/diagram:
80GB + 80GB = 80GB (72.8GB in reality)80GB + 160GB = 80GB (same as above)80GB + 80GB + 80GB = 160GB (147GB in reality)80GB + 80GB + 1TB = 160GB (same as above)80GB + 300GB + 1TB = 380GB (352.9GB in reality)1TB + 1TB + 1TB + 1TB = 3TB (2.7TB in reality )
It’s worth mentioning that I took those figures directly from their simulation engine. But notice that the scenarios I generated follow that diagram I made.1
For those of you who aren’t aware of how RAID systems work, in context of Drobo, here’s all you need to know:
Drobo (and similarly configured RAID arrays) will make sure that if any one of your drives fail, you have enough space to have data parity to complete the file from the remaining drives. This means that data in any given drive is redundantly available in the other drives - therefore reconstruction of lost data is possible in the event of a drive failure.
Naturally, in scenarios 2, 4, 5 and 6, if Drobo allowed you to use the full capacity of the spanned drives, then you suddenly lost your biggest drive (or any drives for that matter), obviously the drive(s) left can’t possibly generate space and data out of thin air.
Instead, Drobo always reserves space for data redundancy and only makes available to you space that can be protected.
The only caveat here is that just like in a RAID 5 array, it can protect you from a single drive failure at a time - this means that it has to be in a healthy (safe) state before you can afford to lose a drive. I have yet to confirm what will happen to your data if two or more drives fail at the same time.
Going Beyond Safety
The only worrying thing about this gadget is its most crucial point of failure: the Drobo itself.
The Drobo is the only thing that can make sense of Drobo formatted disks - so if the unit itself dies, I’m pretty sure that you’re down the shit-river without an oar until you get another Drobo. For all its worth, if all your drives are still healthy, plugging everything into a new Drobo will result in all your data being intact.
What would be nice (but I’m not holding my breath) is that Data Robotics Inc. would give users software that could read/recover data off of Drobo formatted disks. The software need only be able to read/recover data from the drives; I even don’t mind if I have to endure the whole “insert hard drive x” every minute - for as long as I can recover data outside of the Drobo in an emergency.
- Only caveat is that when you ADD disks in the green area, the greater than or equal to sign applies to any ONE of those drives compared to the largest drive.
↑
post updated on July 31, 2008 @ 12:49PM
