Solid State SSD Review for Corporate Users Kingston DC500R

Original author: StorageReview Enterprise Lab
  • Transfer
Kingston recently released the Kingston DC500R enterprise-class solid state drive , designed for high constant loads. Now many journalists are actively testing the new product and release interesting materials. We want to share with Habr one of the detailed reviews of Kingston DC500R, the testing of which will please readers. The original is on the Storagereview website and published in English. For your convenience, we have translated the material into Russian and place it under the cut. Enjoy reading!

Drives Kingston DC500RBased on 3D TLC NAND flash memory technology. Options are available in capacities of 480 GB, 960 GB, 1.92 TB and 3.84 TB, which provides an additional choice for companies that would like to save money, or for those who simply do not need high-capacity drives. This review discusses a 3.48 TB version with declared sequential read and write speeds of 555 MB / s and 520 MB / s, respectively, and 4 KB block read and write speeds at constant loads of 98,000 and 28,000 input operations -Output per second (IOPS), respectively. As part of this product family, Kingston also offers the DC500M, optimized for mixed-use applications.

Kingston DC500R Specifications


For testing enterprise-class solid-state drives with real applications, Lenovo ThinkSystem SR850 was used , and for synthetic testing, Dell PowerEdge R740xd . The ThinkSystem SR850 is an optimized quad-core platform that delivers processor computing power that far exceeds that required for testing high-performance local storage. For synthetic tests, in which CPU capabilities are not so important, a more traditional server with two processors was used. In both cases, we hoped to get local storage performance at the level declared by the manufacturer.

Lenovo ThinkSystem SR850

  • 4 Intel Platinum 8160 processors (2.1 GHz, 24 cores)
  • 16 DDR4 ECC DRAM with 2666 MHz 32 GB each
  • 2 RAID 930-8i 12 Gb / s adapters
  • 8 NVMe drives
  • VMware ESXI 6.5 Software

Dell PowerEdge R740xd

  • 2 Intel Gold 6130 processors (2.1 GHz, 16 cores)
  • 4 DDR4 ECC DRAM memory modules with a frequency of 2666 MHz in a volume of 16 GB
  • RAID PERC 730 12 Gb / s, 2 GB buffer
  • NVMe Embedded Adapter
  • OS Ubuntu-16.04.3-desktop-amd64

Testing information

Enterprise-level test lab StorageReview provides ample opportunities for testing storage devices in an environment that is close to real conditions. The laboratory includes various servers, network devices, power systems, and other network infrastructure. This allows our employees to create close to real conditions for an accurate assessment of equipment performance.
Information about the environment and protocols is included in the reviews so that IT professionals and those responsible for the purchase of storage systems can evaluate the conditions under which the corresponding results were achieved. Manufacturers of test equipment do not pay for the review and do not control it.

Application Workload Analysis

To properly evaluate the performance parameters of enterprise-class storage devices, it is important to model the infrastructure and application workloads that are appropriate for real-world environments. Therefore, to evaluate the Samsung 883 DCT SSDs, we measured the performance of the MySQL OLTP database using the SysBench utility and the performance of the Microsoft SQL Server OLTP database by emulating the TCP-C workload. In this case, for applications, each drive will process from 2 to 4 equally configured virtual machines.

SQL Server Performance

Two virtual disks are configured for each SQL Server virtual machine: a boot disk with a capacity of 100 GB, as well as a disk with a capacity of 500 GB intended for hosting the database and log files. In terms of system resources, each virtual machine was equipped with 16 virtual processors, 64 GB of DRAM, and a LSI Logic SAS SCSI controller. Previously, with Sysbench workloads, we tested I / O speed and drive capacity utilization. SQL tests, in turn, help evaluate latency.

As part of the test, SQL Server 2014 is deployed to guest virtual machines running Windows Server 2012 R2. Loads are created using Quest Benchmark Factory for Databases software.StorageReview's Microsoft SQL Server OLTP Database Testing Protocol uses the current version of the Transaction Processing Performance Council Benchmark C Performance Assessment Program (TPC-C). This real-time transaction processing performance evaluation program simulates the processes of complex application environments. TPC-C testing allows you to more accurately determine the strengths and weaknesses of the storage infrastructure in database environments, rather than artificial performance testing. As part of our tests, each instance of the SQL Server virtual machine worked with a 333 GB (1500 scale) SQL Server database. Performance and latency measurements during transaction processing were carried out under a load of 15,000 virtual users.

SQL Server test configuration (for each VM):
• Windows Server 2012 R2
• Disk space: 600 GB allocated, 500 GB used
• SQL Server 2014
- Database size: 1,500 scale
- Number of virtual clients: 15,000
- RAM memory buffer: 48 GB
• Test duration: 3 hours
- 2.5 hours - preliminary stage
- 30 minutes - direct testing

According to the results of evaluating the performance of SQL Server transaction processing, the Kingston DC500R drive was only slightly behind the Samsung 883 DCT, showing a total performance of 6290, 6 transactions per second (TPS).

Even better than the TPS metric, evaluating the performance of SQL Server is a measure of latency. Here, both drives - Samsung 860 DCT and Kingston DC500R - showed the same time: 26.5 ms.

Performance with Sysbench

The next test used the Percona MySQL database . OLTP performance was evaluated using the SysBench utility. In this case, the average TPS and delay are measured, as well as the average delay time in the most unfavorable scenario.

Every Sysbench Virtual Machineused three virtual disks: a boot disk with a capacity of about 92 GB, a disk with a pre-installed base of this volume of about 447 GB, and a disk with a test database with a capacity of 270 GB. In terms of system resources, each virtual machine was equipped with 16 virtual processors, 60 GB of DRAM and a LSI Logic SAS SCSI controller.

Sysbench test configuration (for each VM):

• CentOS 6.3 64-bit
• Percona XtraDB 5.5.30-rel30.1
- Number of database tables: 100
- Database size: 10 000 000
- Number of database threads: 32
- RAM memory buffer: 24 GB
• Test duration: 3 hours
- 2 hours - preliminary stage, 32 threads
- 1 hour - direct testing, 32 threads

Based on the Sysbench transaction processing performance rating, the DC500R was behind the competition, delivering performance of 1,680.47 transactions per second.

In terms of average latency, the DC500R also ranked last in the ranking with a figure of 76.2 ms.

Finally, after testing the delay time in the worst case scenario (99th percentile), the DC500R was again at the end of the list with a result of 134.9 ms.

VDBench workload analysis

When testing storage devices, application testing is more preferable than synthetic tests. However, although their results do not correspond to real conditions, synthetic tests due to the repeatability of tasks are convenient for determining basic indicators and comparing competing solutions. These tests offer a wide range of profiles, from four corners tests and sample database migration tests to tracking captures from various VDI environments. In all these cases, a single vdBench workload generator with a script processor is used to automate and collect the results in a large cluster of computational tests. This makes it possible to use the same workload for a wide range of drives, including flash arrays and individual drives. As part of the test, we completely filled the drives with data, and then divided them into partitions with a capacity of 25% of the original, in order to simulate application loads and evaluate the behavior of the drive. This approach differs from fully entropy tests, in which, under constant loads, the entire disk is used immediately. For this reason, the following results reflect more stable recording speeds.

• Random read 4 KB: read only, 128 threads, I / O speed from 0 to 120%
• Random write 4 KB: write only, 64 threads, I / O speed from 0 to 120%
• Sequential read 64 KB: read-only, 128 threads, I / O speed from 0 to 120%
• Sequential write 64 KB: write only, 64 threads, I / O speed from 0 to 120%
• Synthetic databases: SQL and Oracle
• Copy VDI (full copy and creating linked copies)

In the first VDBench workload test (4K random read), the Kingston DC500R showed impressive results: a delay of 1 ms until reaching a speed level of 80,000 IOPS and a maximum speed of 80,209 IOPS with a delay of 1.59 ms.

All the drives tested showed almost the same results in the second test (Random Write 4 KB): the speed is slightly higher than 63,000 IOPS with a delay of 2 ms.

Moving on to sequential loads, we first looked at 64K reads. In this case, the Kingston drive maintained a delay within a millisecond until it reached 5200 IOPS (325 MB / s). The maximum figure of 7183 IOPS (449 MB / s) with a delay of 2.22 ms brought this disk to second place in the overall standings.

When testing sequential recording operations, the Kingston device outperformed all competitors by holding the delay below 1 ms up to a speed of 5700 IOPS (356 MB / s). The maximum speed was 6291 IOPS (395 MB / s) with a delay of 2.51 ms.

After that, we moved on to SQL tasks, where the Kingston DC500R was the only device whose delay level went beyond the millisecond in all three tests. In the first case, the disk showed a maximum speed of 26411 IOPS with a delay of 1.2 ms.

In the SQL 90-10 test, Kingston took the last place with a maximum speed of 27339 IOPS with a delay of 1.17 ms.

The same thing happened in the SQL 80-20 test. The Kingston device in this case showed a maximum speed of 29576 IOPS with a delay of 1.08 ms.

Oracle workload test results put DC500R back in last place, but in two tests the device still showed a delay within a millisecond. In the first case, Kingston’s maximum disk speed was 29098 IOPS with a delay of 1.18 ms.

In the second test (Oracle 90-10), the DC500R achieved 24555 IOPS with a delay of 894.3 μs.

In the third test (Oracle 80-20), the maximum speed of the device from Kingston was 26401 IOPS with a delay level of 831.9 μs.

Then we moved on to copying VDI — creating full and linked copies. In testing the download of a full copy of VDI, the Kingston drive again failed to outperform its competitors. The device supported a delay below 1 ms up to a speed of about 12000 IOPS, and the maximum speed was 16203 IOPS with a delay of 2.14 ms.

When testing the Initial Login copy of the VDI, the Kingston device performed better, eventually taking (with a slight margin) the second place. The drive supported a delay within a millisecond until a speed of 11000 IOPS was reached, and the maximum speed was 13652 IOPS with a delay of 2.18 ms.

Also with a slight margin, the Kingston drive took second place on Monday Login test results for a full copy of VDI. The Seagate Nytro 1351 drive showed slightly higher maximum speed, but the Kingston device generally showed lower latency levels throughout the test. The maximum speed of the DC500R was 11897 IOPS with a delay of 1.31 ms.

In testing downloads of linked copies of VDI, Kingston was in last place. The delay went beyond 1 ms already at speeds of less than 6000 IOPS. The maximum speed of the DC500R was 7861 IOPS with a delay of 2.03 ms.

However, according to the results of the Initial Login test, the drive again took second place: the delay exceeded the limits of a millisecond only after reaching peak performance, which ultimately amounted to 7950 IOPS with a delay of 1.001 ms.

In the last test of the linked copy of VDI - Monday Login - the disk also showed a second result: the maximum speed at 9205 IOPS with a delay of 1.72 ms. Outside the millisecond, the delay exceeded when the speed of 6400 IOPS was reached.


The DC500R is Kingston’s latest solid state drive for enterprise users. The DC500R is available in a 2.5-inch form factor. Available capacities ranging from 480 GB to 3.84 TB. The drive is based on 3D TLC NAND flash memory technology and combines a large resource and a high level of performance. For a 3.48 TB disk, sequential read and write speeds of 555 and 520 MB / s, respectively, read and write speeds with constant loads of 98,000 and 28,000 IOPS, respectively, and a resource capacity of 3504 TBW are claimed.

To evaluate the performance of the Kingston DC500R, we compared it with other popular SATA SSDs, including the Samsung 860 DCT and 883 DCT drivesas well as a Seagate Nytro 3530 drive . Kington DC500R was able to stay at the level of competitors, and in some cases even surpass them. When testing application workloads, Kingston DC500R performed quite well when processing SQL tasks, taking the overall second place in terms of transactions per second (6291.8 TPS) and latency (26.5 ms). In a Sysbench test with more intense write performance, the DC500R ended up at the end of the list with 1,680.5 TPS performance with an average latency of 76.2 ms and a latency of the least favorable scenario at 134.9 ms.

As part of testing random read and write blocks of 4 KB, Kingston DC500R demonstrated a speed of 80209 IOPS and a delay of 1.59 ms when reading, as well as a speed of 63000 IOPS and a delay of 2 ms when writing. As part of a 64 KB read and write test, the DC500R reached speeds of 7183 IOPS (449 MB / s) with a delay of 2.22 ms and 6291 IOPS (395 MB / s) with a delay of 2.51 ms, respectively. In the context of synthetic tests using SQL and Oracle databases and increased write speed requirements, the performance of the DC500R was poor. In terms of SQL workloads, Kingston DC500R took the last place in all three tests and turned out to be the only drive whose delay level went beyond the millisecond. However, in testing Oracle the picture was much better. In two out of three tests, the drive supported a delay of less than 1 ms, which took it second place. Kingston DC500R showed a decent level of performance when testing using copies of VDI - both full and related.

Overall Kingston DC500R SSD- A high-quality device in its class that deserves closer attention. No matter how much we like high-performance technologies (NVMe and analogs), it is SATA drives that remain the most preferable solution for processing tasks where reliability plays a crucial role, for example, loading a server or storage controller. Such drives are also an economical solution for storing server data in situations where value for money is of great importance. In addition, they offer all the benefits in terms of total cost of ownership that solid state drives stand out from the background of hard drives. The DC500R performance allows the device to occupy the top lines according to the results of many tests conducted by us in comparison with other worthy drives.

DC500 Series models are available from authorized Kingston distributors.
For questions of testing and validation, you can contact the Kingston Technology representative office in Russia by e-mail at

For more information about Kingston Technology products, visit the company's website.

Also popular now: