Drobo Mini review: An underperforming and overpriced storage device

But there's no benefit to picking the highest amount, either, since each of the Mini's four drive bays supports a drive of 1TB at most. For that reason, it's kind of pointless to select a volume that's larger than 4TB. Having a "virtual" volume that's larger than the actual storage space can also be confusing since you end up thinking you have more storage that you actually do. Our review unit, with four 750GB hard drives, actually offers about 2TB of protected storage space with one-disk redundancy, or 1.3TB with dual-disk redundancy.

The formatting process required the computer to restart one more time before the drive is usable. So all in all, if everything goes well, you'll spend about 5 minutes downloading and installing the Drobo Dashboard software, another 5 minutes on formatting the drive, and have to restart restart the computer twice, to complete the setup process for the Drobo Mini. This isn't really that long of a process, unless you compare it with setting up any other Thunderbolt storage device.

Technically, it seems you can just ignore the Drobo Dashboard software and use Disk Utility to format the Mini the way you would with other drives. But doing that would mean you can't take advantage of BeyondRAID, which is the strong point of the drive and is the only option the Drobo Dashboard software gives you. You can't use the software to set up the Drobo Mini using a standard RAID setup, such as RAID 0, RAID 1, RAID 10, or RAID 5. (For more about RAID, check out this post .)

In my trials, BeyondRAID worked as intended, though I didn't get the chance to test all it could do. This is because I didn't have any internal drive larger than 750GB on hand at the time of testing. In fact, there aren't many 1TB 2.5-inch drives on the market that spin at 7,200rpm. Nonetheless I was able to remove one of the hard drives and replace it with another of the same capacity without having to power down the Drobo Mini, something RAID 5 or RAID 1 storage devices also offer. The Drobo Mini took about 20 minutes to add a new drive in my trials with about 60GB of storage space being used. Note that the RAID rebuild time varies a great deal depending on the amount of existing data, the more data, the longer it takes. During this rebuild time all the status light were going wild, continuously flashing different colors, and the Drobo Dashboard software gave me warning instructions. In the end the operation was a success. The Drobo Mini is the only Thunderbolt storage device that offers the option of scaling up its storage space dynamically this way. With other multiple-drive Thunderbolt storage products, such as the Pegasus R6 or the WD VelociRaptor Duo, you can hot-swap existing drives but can't scale up the drive's storage space.

Performance
The Drobo Mini isn't a superslow external storage device, but overall it's the slowest among all multiple-bay Thunderbolt storage devices I've reviewed.

To be fair, it's the only multiple-bay Thunderbolt storage device that uses 2.5-inch (laptop) hard drives, while the rest use the larger 3.5-inch hard drives, which tend to be faster. On the other hand, however, the Drobo Mini was tested with four high-end Seagate Momentus 750GB hard drives (model ST9750420AS) that spin at 7,200rpm and are faster than some 3.5-inch hard dives used in other Thunderbolt storage solutions. The WD My Book Thunderbolt Duo, for example, uses two 3.5-inch WD Green hard drives -- designed with more focus on energy conservation than speed -- that are likely slower than the Seagate Momentus, individually.

The Drobo Mini uses 2.5-inch hard drives, which might or might not be the reason why its performance is not on par with 3.5-inch-based Thunderbolt storage devices.
The Drobo Mini uses 2.5-inch hard drives, which might or might not be the reason why its performance is not on par with that of 3.5-inch-based Thunderbolt storage devices. Dong Ngo

Every internal drive on the market, be it a hard drive or an SSD (which caps by the speed of SATA 3 -- 6Gbps -- at most), by itself is much slower than the Thunderbolt standard, which caps at 10Gbps. That said, many multiple-bay Thunderbolt storage devices, such as the Pegasus R6,or the WD VelociRaptor Duo, are able to aggregate the speed of individual drives into a faster combined speed. I expected similar performance from the Drobo Mini, especially after all the hype Drobo has for it in terms of performance, but it didn't really deliver.

I tested the drive both against other Thunderbolt storage solutions (and internal drives) and against other external drives, including USB 3.0 drives. The test machine is a late-model 2011 MacBook Pro, running a high-end SATA 3 SSD. The Drobo Mini was tested only in BeyondRAID with one disk redundancy, which is similar to RAID 5, since this is the default option and generally offers faster performance than the dual-disk redundancy, similar to RAID 1. All of the tests were performed by timing with a stopwatch how long the drive took to finish a heavy data transfer job. The scores in all of CNET's storage reviews are those of real-world sustained speeds. Let's check how the Drobo Mini performed.

Tests against other Thunderbolt devices and internal drives
These are the test where the Drobo Mini was stacked against other Thunderbolt storage devices and internal drives. The reason internal drives (mostly SSDs) are thrown in to the mix is because prior to Thunderbolt, they are the fastest storage devices on the market, so it's just natural to use them to show how much faster Thunderbolt can be. Like all other Thunderbolt storage devices, the Drobo Mini had to go through two rounds of testing.

In the first, called "Self read and write," I copied data within the drive, from one folder to another. In this test it scored just 45MBps. In this same test, the WD My Book Thunderbolt Duo scored 77MBps and 40MBps in RAID 0 and RAID 1, respectively. The Pegasus R4, a RAID 5, four-bay drive, scored 101MBps in this test.

In the second test, call "Unit to unit," the Drobo Mini was set to copy data from another Thunderbolt storage device. This test is to show the drive's best possible write performance. I picked the high-end WD VelociRaptor Duo, formatted in RAID 0, as the source of the data to make sure the source is not the bottle neck. In this test, the Drobo Mini scored 101MBps, while the WD My Book Thunderbolt Duo scored 101MBps and 168MBps in RAID 0 and RAID 1, respectively. The R4 scored 240MBps in this test.

Thunderbolt vs. Internal hard-drive performance (in MBps)
(Longer bars indicate better performance)
Unit to unit   
Self read and write   
Promise Pegasus R6 (RAID 5)
323.79 
192.53 
Plextor PX-256M2S
261 
162.03 
OCZ Vertex 3
260.71 
150.01 
Crucial M4
235.51 
117.99 
LaCie Little Big Disk SSD
233.5 
141.69 
OCZ Agility 3
207.75 
101.67 
Patriot Wildfire
202 
99.72 
WD VelociRaptor 600GB
126.33 
58.05 
Elgato Thunderbolt SSD
121.96 
71.84 
Seagate Barracuda XT
115.71 
51.1 
WD VelociRaptor 300GB
112.59 
47.12 
Drobo Mini
101.17 
45.16 

Note on the chart how the Drobo Mini was significantly slower than most others. In fact, it's by far the slowest non-RAID1 Thunderbolt RAID system to date. Generally, Thunderbolt storage devices with RAID 0 or RAID 5 (which is similar to how the Mini's BeyondRAID was set up for the test), are at least on par or faster than internal SSDs, but the Drobo Mini couldn't manage to outdo the WD VelociRaptor regular hard drive in terms of data rate.

Tests against external drives

In this set of tests, the Drobo Mini was tested the way it's most likely to be used in the real world: transferring data back and forth between itself and a computer. For this test, when using the Thunderbolt connection, the Mini scored 106MBps for writing, making it almost the slowest among Thunderbolt storage devices, including single-volume ones. In reading, it did even worst at 95MBps, clearly the slowest I've seen. Generally, multiple-bay storage devices that are not using RAID-1 are faster than single-volume storage device, in the case of the Drobo, it could barely keep up with even the Buffalo MiniStation HD-PATU3, which is a Thunderbolt portable drive that's based on a single internal 2.5-inch hard drive.

When used with a USB 3.0 connection, the Drobo Mini was much slower, at 59MBps and 77MBps for writing and reading, respectively, making it the slowest USB 3.0 storage device to date.

Data transfer (TB vs. external, in MBps)
(Longer bars indicate better performance)
Read  
Write  
Promise Pegasus R6 (RAID 5)
177.53 
210.5 
LaCie Little Big Disk SSD
186.8 
184.71 
Promise Pegasus R4 (RAID 5)
171.1 
150.47 
LaCie Rugged Thunderbolt
154.9 
150.07 
Elgato Thunderbolt SSD  
168.97 
120.61 
Drobo Mini (via Thunderbolt)
94.66 
106.03 
Drobo Mini (via USB 3.0)
76.79 
59.26 

All in all, in terms of data speed, the Drobo Mini is unimpressive and falls to a new performance low for both Thunderbolt and USB 3.0 storage devices. I also noticed something troubling: the drive seemed to become progressively slower the more data it stored. For example, when the drive was 1.5TB full, out of its 2TB protected storage space, it scored just 85MBps for writing when copying data from the MacBook Pro test machine, though it managed 106MBps at this job earlier when it was empty. Judging from the performance alone, I can't help but feel there's something not right about the BeyondRAID itself. Unfortunately there was no option to make the Drobo Mini work in any of the standard RAIDs.

I also noticed that the Drobo Mini became hot very fast and remained hot during operation. It wasn't hot enough to cook an egg on, but it was definitely the hottest of the Thunderbolt storage devices, despite those loud ventilation fans.

Conclusion
Unfortunately, the Drobo Mini over-complicates external storage. It offers very little in real-world benefits and in the end, users pay too much for the simple privilege of owning something that's different. There are four important factors when it comes to getting an advanced storage system: performance, ease of use, price, and data protection. While in theory the Drobo Mini might offer a relatively new way to protect and scale up your data, it categorically and miserably fails on the rest, making it not only a waste of money but also a waste of time.

Editors' Top Picks

 

ARTICLE DISCUSSION

Conversation powered by Livefyre

About The Author

CNET editor Dong Ngo has been involved with technology since 2000, starting with testing gadgets and writing code for CNET Labs' benchmarks. He now manages CNET San Francisco Labs, reviews networking and storage products, and also writes about other topics from online security to new gadgets and how technology impacts the life of people around the world.