Archive Home arrow Reviews: arrow Network arrow QNAP TS-879U-RP NAS Network Storage Rack Server

QNAP TS-879U-RP NAS Network Storage Rack Server E-mail
Reviews - Featured Reviews: Network
Written by Bruce Normann   
Wednesday, 08 February 2012
Table of Contents: Page Index
QNAP TS-879U-RP NAS Network Storage Rack Server
QNAP v3.5 New Features-Home
QNAP v3.5 New Features-Business
Closer Look: QNAP TS-879U-RP
Insider Details: QNAP TS-879U-RP
Technology Details: QNAP TS-879U-RP
QNAP Turbo NAS Features
QNAP TS-879U-RP NAS Hardware
QNAP TS-879U-RP Software
QPKG Center Software Expansion
NAS Testing Methodology
Basic-Disk Test Results
RAID 5 Test Results
NAS System Overhead Measurements
NAS Server Final Thoughts
QNAP TS-879U-RP Conclusion

1GB RAID 5 Test Results

If you've got more than three HDD spindle to put in play, it makes sense to use one of the more advanced RAID configurations. RAID 5 is one of the most popular setups, primarily due to the balance it exhibits between capacity and redundancy. Not surprisingly, most NAS units that can support more than three HDDs also support RAID 5, so it makes sense to use it for test purposes. Most NAS products that can support RAID 5 go beyond the minimum number of drive bays, to a total of four, so that is the number of drives that I typically use to test with, even though I could get by with only three. I also took advantage of the massive capacity that the QNAP TS-879U-RP offers, and tested it with the full complement of eight drives. I am fortunate to have purchased eight W-D Caviar Black 750GB drives before the floods in Thailand wiped out half the world's HDD production capacity.

The results for RAID 5 read testing are very similar to single disk testing, which is not a bad thing. Given all the behind-the-scenes processing that goes on to calculate parity bits, these results show that most QNAP Turbo NAS units have the necessary power under the hood to keep the drives performing at their highest potential during read operations. When using RAID 5, the TS-879U-RP outperformed all of the other NAS systems and came close to its own performance level in single-disk operation. There was virtually no difference between 4-disk and 8-disk performance. CPU, memory, and network utilization were almost identical too, which tells me the support hardware has plenty of muscle to handle the task of pushing around all eight disks. Read performance is clearly very strong with this system, which is a real bonus if you use it as frontline storage. Using it primarily as a backup system, you want top-notch write performance, which we'll test next.


The 1 GB RAID 5 disk write test stays right up there with the single disk results. It's well known that RAID 5 write performance can be a weak point, with all the computation overhead involved and the extra parity bits that need to be calculated and written to each of the drives. The only way to overcome that is with raw computational horsepower, which is what the TS-879U-RP brings to the table better than any other NAS in this test. It's a shame that the simplest task any NAS can perform is basic backup duty, and in order to do that well, you need to buy the most powerful system to effectively reap the benefits of a multi-disk array. The reduced write performance with 1500 MTU is also a factor in RAID 5, so it's looking more and more like Jumbo Frames (9000 MTU) should be the preferred network setting for this unit. Unfortunately, you can't take advantage of Link Aggregation at that setting, as it is limited to 1500 MTU.


Next up is 10 GB (1000 metric megabytes / 10,000,000,000 bytes) file transfer testing. Using the 4-disk RAID 5 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. I'll also include some 8-disk results, just because I can.

10GB RAID 5 Test Results

Looking at read tests with a single 10GB file, the TS-879U-RP still beats out all of its little brothers, but not as much as in some of the other tests. The TS-659 Pro II improved its performance when transferring large files; it shows how this unit is built to carry the heavier data loads. The TS-879U-RP takes it a step further and comes in first place as a result. The proof of that is the nearly identical performance between 4-disk and 8-disk operation. There is still a small decrease in performance with Jumbo Frames disabled, which is now consistent between Read and Write tests, as we'll see in a moment. It's not significant enough to outweigh the reliability and availability advantages of Link Aggregation, though. Not in a corporate LAN room, or anywhere else.


Looking at write tests with a single 10GB file, the TS-879U-RP smokes the smaller units with gains of 50% and 100% over the Intel Atom and Marvell-based models. If you're writing large files to a NAS, you can't afford to scrimp on system hardware; you need the biggest, baddest CPU you can afford. The other units still suffers from the typical RAID 5 write penalties due to the computing overhead required to deal with the parity bits. The various caches built into the host and target system help out on the smaller file transfers, but they get filled up and lose their effectiveness when dealing with large files like this. Once again, the TS-879U-RP just laughs at the additional load of four additional hard drives; the CPU utilization never got above 25%. Once again, the 1500 MTU results lag a little bit behind the Jumbo Frames performance, but the redundancy of the dual GbE interfaces in a teaming arrangement trumps the small loss in throughput. The bigger issue is that the TS-879U-RP is being completely throttled by the 1000BASE-T interface, and really can't perform at its best without one or two 10GbE connections.


All in all, my impression of the test results is that the QNAP TS-879U-RP puts in a solid performance, eeking the most out of the GbE interface that comes standard. Judging by the System Overhead measurements that I'll show you in the next section, the TS-879U-RP doesn't even break a sweat when its throughput is limited by the network connection, at 1000 megabits/second. It's performance is suited for any task you might think of: front-line storage, backup, replication, or any combination. Its iSCSI capability means you're not limited to a single application, and the VMware capability means it's at home in virtualized environments, as well.

Now, let's take a closer look at the internal workings of the NAS, where we can see the individual activity of the CPU, memory, and network interface. It's these support subsystems that have a big influence on the overall system performance, as I've demonstrated in the past.

NAS Comparison Products

EDITOR'S NOTE: We've retested this product using 10Gb Ethernet, which gave phenomenal results. Read more here: QNAP TS-879U-RP 10GbE NAS Server



# I3 with AES-Ni ?Moogle Stiltzkin 2012-02-16 02:29
Since when did Intel I3's have AES-Ni instruction ?

QNAP currently only has 2 rack models with the x79 name that has XEON cpus which do have AES-Ni.

But the other X79 models only have Intel I3 cpus, and last i checked they didn't have AES-Ni instructions on them ..... yet your saying otherwise ? Got any references for that ?

However despite that, Jason from QNAP is claiming their tests on the 879U-RP with AES encryption could achieve 100MB/s in both read & write in Gigabit environment.

His a trustable guy so i'll believe that :X but what i doubt is your claim that i3 cpus have AES-Ni ....
Report Comment
# Face Palm...Bruce 2012-02-16 06:08
Just looked it up on Intel site, and you're right. There are 141 products with AES-NI. No i3's though.

Well, I've got some re-writing to do....

FWIW, my initial testing with AES-256 supports Jason's claims. The GbE bottleneck is the dominant factor.

Thanks for pointing this out.
Report Comment
# RE: QNAP TS-879U-RP NAS Network Storage Rack ServerMoogle Stiltzkin 2012-02-16 06:35
yeah, all that matters is the result. doesn't matter if it has AES-Ni or not (although it's good to have since it's suppose to accelerate AES encryption so why not) since all comes down to whether it can perform well even with AES 256 encryption enabled.

Jason says it does 100 mb/s and that in my opinion is very good. I was worried and confused why they opted for the i3's which are basically almost the same as i5's but without the aes-ni. But from the results, seems my worries were unfounded ? As i was expecting results like these which was an article by tomshardware showing how bad performance was on many NAS brands across the board that didn't use encryption acceleration such as AES-Ni.,2873.html

Anyway i wonder if you would be kind enough to add to your review, a chart showing AES 256 encryption performance of the i3 QNAP to help back up Jason's claim. I trust Jason, but other people who don't know him will want proof from third party sites like yours to see if what QNAP claims is true or not.

I'm also interested to know whether the other i3 QNAP models such as the 1079 can achieve the same good results for aes 256 encryption, as well as compare the results with Synology Diskstation DS3611xs which sports a Intel® Xeon® Processor E3-1225 cpu.
Report Comment
# Follow-Up testingBruce Bruce 2012-02-16 07:07
Yes, I plan to do a follow-up article which will focus on performance improvements with the optional 10GbE interface, and to document the AES-256 performance. Take a look at the last chart in the NAS System Overhead Measurements section of this article. That's one of the tests I ran with AES-256 enabled, and the CPU looks like it still has some headroom left, when running of the Gigabit interface. I'll have full results in the follow-up article.
Report Comment
# RE: QNAP TS-879U-RP NAS Network Storage Rack ServerMoogle Stiltzkin 2012-02-16 06:44
FWIW, my initial testing with AES-256 supports Jason's claims. The GbE bottleneck is the dominant factor.

And lets not forget conventional hard drives may also be a bottle neck ?

The fastest read and write maximum throughput for a hard drive is Seagate's New Barracuda 3TB (ST3000DM001)

Read Throughput Maximum: h2benchw 3.16

Write Throughput Maximum: h2benchw 3.16'
Report Comment
# 8xBruce Bruce 2012-02-16 07:12
I'll be running RAID 5 with eight drives, though.

I don't expect to reach the same speeds that QNAP did with (8) Intel SSDs, but I'm betting it will still be way above what the unit does with the Gigabit interface.
Report Comment
# RE: QNAP TS-879U-RP NAS Network Storage Rack ServerMoogle Stiltzkin 2012-02-16 07:11
regarding your last con, well some desktop drives can be used fine for raid.
For me, i've used 6 x SAMSUNG HD203WI for a long time and they work fine on my QNAP ts-659 pro II.

QNAP and other branded NAS, tend to ignore tler, so it's not critical for these nases when using raid seeing as it's not being used.

The responses I received from Synology, QNAP, NETGEAR and Buffalo all indicated that their NAS RAID controllers don't depend on or even listen to TLER, CCTL, ERC or any other similar error recovery signal from their drives. Instead, their software RAID controllers have their own criteria for drive timeouts, retries and when a drive is finally marked bad.

however samsung has sold off their hard drive business to Seagate who bought their hd unit out.

So the only other manufacturer that offered some good desktop raid drives was Hitachi i believe which was the Hitachi Deskstar 7K3000 3TB which is a desktop drive that works fine in raid for say a QNAP.

The desktop drives you should be wary about for using raid are western digital and seagate, which intentionally make it problematic in a raid setup, by dropping out very often to make you buy their x2 expensive raid edition drives.
Report Comment
# Some work fine.....Bruce Bruce 2012-02-16 07:44
The Spinpoint F3 was a favorite for NAS use, no doubt. Other readers have also reported good luck with their Hitachi Deskstars. You'll see similar reports on the QNAP support forums. I'm somewhat concerned that most reports are from users with smaller NAS units, where the combined vibration of 8-12 drives is not present. It's bearing failure, as much as controller quirks that cause HDD failures.

Samsung drives are still being built to the old designs/specs for now, right? Get them while you can, I think....
Report Comment
# The 20 should be closser to 2. bits vs Bytes.tygrus 2012-02-22 03:05
"twenty times faster than what the latest generation of SSDs can muster"
Sorry but 10GbE =
Report Comment
# Comment system ate my commenttygrus 2012-02-22 03:41
There was some mathematical symbols in my previous comment. The comment has been truncated.
10GbE =lt 1250MBps. 500MBps = 4800Mbps.
Reasonable sound level for a device to be in a server room.
Too loud for home use.
Would like to see aditional testing with multiple clients or larger queue depth. Need to beg someone for some 10GbE hardware.
Report Comment
# 10GbE is comingBruce 2012-02-22 06:00
Next week I'll have two Intel 10GbE NICs. I'll just have the one PC, but would RoboCopy, with it's multithreaded operation get closer to the multiple client scenario?
Report Comment
# Updated Results in New ArticleBruce 2012-03-21 08:00
Just wanted to let all of you know that I completed the additional testing, with 10GbE NICs and a RAM Disk on the PC.

Wow! What a difference.

Check out the results in my follow-up article here:
Report Comment
# RE: QNAP TS-879U-RP NAS Network Storage Rack ServerMoogle Stiltzkin 2012-03-21 08:09
Nice :} just read it.
Report Comment

Comments have been disabled by the administrator.

Search Benchmark Reviews

Like Benchmark Reviews on FacebookFollow Benchmark Reviews on Twitter