|IOCELL NetDISK 351UNE Network Storage Device|
|Reviews - Featured Reviews: Network|
|Written by Bruce Normann|
|Monday, 28 November 2011|
Page 8 of 11
1GB Single-Disk Test Results
The bottom line for any storage device is the combination of capacity and transfer speed. For a network attached storage server, the differences are all about the infrastructure that is placed around the basic HDD array. Since capacity is something that's easy to define and measure, the real question for any NAS product is how fast will it Read and Write data. For this reason, Benchmark Reviews measures NAS performance as the bandwidth achieved during a file transfer test. The first tests we perform utilize a single 1GB (1000 megabytes / 1,000,000,000 bytes) file in a transfer to and from the NAS.
Since we started testing NAS units exclusively with Win 7, there aren't as many prior test results to compare with. We'll try and build up the NAS testing as best we can in the next year. In the meantime, we can compare the IOCELL NetDISK product to several units from QNAP that we have on hand now. In many ways, it's not a fair comparison, given the price difference between the IOCELL NDAS unit, and all of the QNAP products. There's like a 5X difference between the NetDISK 351UNE and the lowest-priced QNAP NAS Server in this test.
With all the NAS units operating in single disk mode, the 351UNE is not at the same level as the high-end NAS units, but it's not embarrassed. The average read speed was 62 MB/s, which is about 75% of the average performance that the QNAP units achieved. Not bad, for one fifth the price! Of course, none of them can compare with the performance of an SSD and a Velociraptor attached directly to the Southbridge.
Moving on to the 1 GB write bandwidth test, the performance of the NetDISK 351UNE drops a fraction, reaching an average write speed of 54 MB/s. In all cases, I detected very minor differences with the NetDISK NDAS, when comparing performance with Jumbo Frames enabled on the host Ethernet controller. Some of the NAS units we have tested had much more sensitivity to the MTU value in use, as you can see from the chart below. The driver for the NDAS device did not allow access to the MTU settings of the NetDISK unit, so the results may have been different if we were able to configure the IOCELL hardware directly. The NAS results here mirror the cost and performance of the CPU included in each device, and clearly the speed goes up with price. The NDAS does it all without a CPU, and it basically matches the performance of the lowest price NAS, which is running a 1.6 GHz Marvell ARM-based CPU.
Next up is 10 GB (1000 metric megabytes / 10,000,000,000 bytes) file transfer testing. Using the single-disk configuration in each NAS, and a single Gigabit connection, network throughput will be put to the test, and the effect of any system or hardware caches will be minimized.
10GB Single-Disk Test Results
Examining 10GB basic file transfer speeds, the NetDISK 351UNE only loses about 1 MB/s compared to the 1 GB results. The NAS units gained some, with the use of this larger test file, such that the NDAS only reaches approximately 64% of the performance of the average of the four NAS devices, compared to the 75% we got with the smaller test file. Still, I'm impressed with the results of the lower-cost NDAS device, and it is clearly performing in the range of its advertised ratings. Once again, the Jumbo Frame/MTU value had very little impact on the results for the NetDISK product, most likely due to the fact that the MTU settings are not configurable on the 351UNE model.
In our 10GB write performance tests, the NetDISK 351UNE actually beats out the lowest priced QNAP NAS by a few MB/s. Putting up a decent result of over 59 MB/s with the larger 10GB test file, it narrowly improved on the 55.9 MB/s performance of the TS-219 P+. That's quite a coup for the NDAS technology, right there!
I wasn't sure what kind of performance I should expect; during the unboxing, assembly and initial setup of the IOCELL NetDISK 351UNE, it all looked too simple and easy to me. But, that's why we test, because first and second impressions don't really matter in this technology sector - performance is almost always the driving factor. On that score, I'm quite impressed, when you consider the cost basis for this product. I know it has very little in the way of software features, and can't do most of the cool things that we've become accustomed to in the world of NAS Servers. For backup purposes and basic networked storage though, it puts in a very respectable performance. The fact that it does so for well under $100 is a minor miracle in my book.
In the next section, we're going to take advantage of the NDAS technology, and the fact that the NetDISK device is configured to look and act just like a local drive. We'll perform some benchmark testing that doesn't normally get put into the mix when we test NAS-type devices.