SAS vs. SATA: Does a Personal Cow Need a Server Saddle?

    Today we decided to talk about hard drives with the SAS interface, and not by chance. A new round of interest in SAS drives came about a year ago with the release of Asus motherboards based on the X58 chipset. Their distinguishing feature was the integrated SAS controller, whereas before, to use this type of hard drives, it was necessary to buy a separate controller, which cost no less than the motherboard. And slowly quite expensive SAS disks with a spindle speed of 15,000 rpm began to move to the top "gaming" personalities. We had the opportunity to evaluate the first test configurations with a RAID array of two SAS Seagate Cheetah 15K.6from large Moscow distributors, and immediately doubts arose about the advisability of installing these server “screws” in a desktop computer, albeit the most expensive one.


    image
    450 Gb SAS Seagate Cheetah 15K.6 Noise, vibration and heat from SAS were immediately noticeable. During the tests, the discs were heated so that it was difficult to hold a hand. But maybe these disadvantages are justified by performance? Indeed, at one time SATA disks with a spindle speed of 10,000 rpm firmly took their place in top-end desktop systems. Well, let's try to answer this question in practice.
    In today's test, we will test 450 Gb SAS Seagate Cheetah 15K.6 ST3450856SS hard drives, both single and in RAID0, and compare their performance with the results of a similar array from SATA's of Seagate Barracuda 7200.12. This will allow us to find out whether hard drives with a spindle speed of 15,000 rpm or not give any advantage.

    Compare the characteristics

    image

    On the Barracuda 7200.12 side, there is more volume, twice as much angular density of data recording, low power consumption as compared with the SAS disks tested in this material, and also twice as much buffer volume. Manufacturers are in no hurry to put a large amount of cache memory on server hard drives - this is due to the fact that storing a large amount of information in the buffer memory reduces the reliability for which these HDDs are designed.

    image

    The Seagate Cheetah server, in turn, has an undeniable advantage in the form of twice the speed and significantly reduced read and write access times.

    image

    Testing technique

    Equipment configuration:
    • Asus P6T6 Revolution Motherboard
    • Intel Core i7 975 Processor
    • Maxtor DiamondMax <6L300S0> Hard Drive for the system
    • RAM 3GB DDR3 1800MHz
    • Microsoft Windows Vista Service Pack SP2


    The following software was used for testing:
    • Everest Ultimate Edition 02/05/1750
    • HAB v0.4a
    • HD Tach 3.0.4.0
    • HD Tune v3.50
    • PCMark04
    • PCMark05
    • PCMark Vantage
    • Iometer-2006.07.27


    When testing single disks, the NTFS file system was used with the default cluster size. Under the partition on the disk, all available space was allocated. For hard drives combined in RAID0, the cluster size was chosen to be 16KB. The maximum possible disk space was allocated under RAID0 from two Seagate Cheetahs, that is, RAID0 from two disks was created on a volume equal to the total volume of disks. The array size from Seagate Barracuda was equal to the size of one disk. RAID0 from disks with SAS-interface was built on the integrated Marvell 88SE6320 controller. SATA drives were tested connected to the ICH10R.
    All tests were performed three times, the graphs show the average values ​​of the obtained HAB data


    image
    “IO Delay” characterizes the time of access to the hard disk, shows the reaction speed of the disk (the time for which the controller issues a request and processes it) and the system (delay in the data transfer channel plus delay in the driver plus delay due to the speed of the processor itself) and speed interface. The smaller this parameter, the better.
    The test ended with the clear advantage of the server disk, its reaction speed is almost two times higher than that of the competitor in this article. It is also worth noting that the access time to an array of two SAS disks is reduced compared to a single disk, in contrast to the same time for SATA disks. I will not venture to assert that this is really possible and is not a mistake of the test program, but the trend itself cannot but rejoice and inspires hope that Seagate Cheetah will be able to please with good speed.

    image
    “Access” is responsible for the average access time, shows how long the head will fly on the surface of the plate until it finds the necessary data. This parameter greatly affects the response of the system - the smaller it is, the better.
    Hopes are justified - the advantage in access time, which affects the subjective sensation of hard drive speed more than read speed, reaches 2.5 times. Both disks gain an advantage in access time when combining in RAID0, but it should be noted that Seagate Barracuda has to do with the fact that the available disk space of the array is only half of the maximum possible disk capacity.

    image
    “Linear read” shows the speed of reading sequentially located data at different block sizes. The higher this speed, the better, of course.
    A SAS drive is faster at any data block size. Its maximum speed exceeds 160 MB / s versus 125 MB / s with Seagate Barracuda, which nevertheless is one of the fastest drives with a rotation speed of 7200 rpm by this indicator. An advantage in the work of the SAS controller is also noteworthy: if Barracuda's RAID0 loses to a single one when working with data blocks smaller than the cluster size, then this does not happen in an array from Cheetah - surely faster operation compared to a single disk.

    Everest, HD Tune, HD Tach.
    The access time of hard drives and arrays of them was also measured in other popular programs.

    image
    Everest


    image
    HD tune


    image
    Hd tach


    It is not surprising that in them a disk with a rotation speed of 15,000 rpm breaks far ahead - 6 ms versus 15 ms. It is interesting that the opinion of various programs on whether the response of a RAID0 array of SAS disks is faster than that of a single disk, divided in half. HD Tach on this issue agrees with the HAB and notes some growth, while the test results of HD Tune and Everest, on the contrary, show a 0.6-0.8 ms deterioration in disk access time.
    Also in these three programs, linear read speeds of hard drives were measured.

    image
    Everest


    image
    HD tune


    image
    Hd tach


    * On the graph with the results in the HD Tach program, there is no Burst Speed ​​value for the array of SAS disks, since the program stably gave out numbers greater than 3500MB / s, which obviously cannot correspond to reality.

    Seagate Cheetah shows simply excellent results for a mechanical disk - even at the end the linear read speed does not fall below 100 MB / s, and in the case of a RAID0-array - below 200 MB / s. In the case of a single disk, the average read speed is 30-50% higher than that of one of the fastest representatives of 7200 rpm of disks. In the case of combining drives in RAID0, the situation on the graphs for SAS-drives is not so advantageous. A feature of this SAS controller is a very smooth linear reading graph for hard disks in the RAID0 array - the differences between the speeds at the beginning and at the end of the disk are quite small compared to the corresponding graph for ordinary hard disks. Hence the seemingly paradoxical loss of server disks in linear read speeds.

    PCMark04
    Next, we compare the performance of hard drives in PCMark packages. Of PCMark04, only one test is interesting - “File Copying”, which is unique, that is, it is found only in this version of PCMark. This test measures the speed of copying a set of files within a single partition of a hard disk.

    image
    PCMark04

    The results of this test (moreover, they repeat well in various operating systems) indicate its unsuitability for testing SAS disks. It simply cannot drive with such a speed of rotation and such a short response time so much to lose to an ordinary hard drive, even if one of the fastest. The acceleration of this test when using RAID0 from server disks was also not observed.

    PCMark05
    The 2005 test package includes the following tests: "Windows XP Startup", which displays the speed of the drive while the operating system is loading; "Application Loading", demonstrating the performance of the disk system when opening and closing six popular applications in succession; "General Usage", which displays the speed of hard drives during the work of a number of frequently encountered applications; "File Write", evaluating the speed of creating files; "Virus Scan", which measures the performance of the hard disk during scanning files in the system for viruses.

    image
    PCMark05 - reading

    The complete fiasco of Seagate Cheetah? No, it’s rather a failure of the Futuremark test. No matter how much I like the graphics benchmarks of this company, its creations are not suitable for testing some hard drives. Of the 10 tests, SAS disks won only two: loading the Windows XP operating system and loading applications in RAID0 array mode. The final result of the HDD Test Suite 2005 sample is presented in the following diagram:

    image
    PCMark05 - final score of the

    PCMark Vantage
    Futuremark's newest overall system performance test includes as many as 8 hard drive performance tests.
    In the “Windows Defender” subtest, the hard drive runs under multithreaded load, one of the streams of which is file scanning. In "Gaming" emulates the behavior of the drive under the load characteristic of computer games. The Windows Photo Gallery subtest evaluates drive performance when downloading images from the photo gallery. Windows Vista Startup emulates the behavior of the drive when the Windows Vista operating system boots. Windows Movie Maker evaluates performance under the load typical of video editing. In the Windows Media Center subtest, the hard drive is tested in the mode typical for working in Media Center. Windows Media Player emulates adding files to Windows Media Player.

    image
    PCMark Vantage - reading

    Judging by the results, little has changed in the PCMark test over the years between versions 2005 and Vantage. SATA disks both individually and in an array look more confident. You can, of course, try to explain this with a large amount of cache memory or a more efficient firmware for Seagate 7200.12, but this advantage does not clearly reflect the true ratio between these disks. A single Cheeatah outperforms Barracuda in a game test, loading Windows Vista. A RAID0 array of server hard drives, in turn, wins when scanning files with antivirus, adding images to the photo gallery, and loading Vista.

    image
    PCMark Vantage - Final Score

    And I do not want to comment. According to the results of PCMark Vantage RAID0 of two 15,000 rpm of hard drives works at the level of a single Seagate Barracuda ...

    Iometer-2006.07.27
    Iometer is a complex fully synthetic test that can simulate the work of a hard disk in various operating modes, for example, as a file server or workstation. Intel IOMeter gives you complete freedom of choice to configure this test application. During testing, IOMeter was configured in accordance with Intel recommendations and the methodology developed by Storagereview.com.
    Intel IOMeter works with so-called workers. For single-processor configurations, Intel recommends creating one such worker. Each worker tests a “target” or “target,” which is either an unpartitioned physical disk or one or more partitions on a disk. For each worker (workers), the so-called work rules are assigned, the “access pattern”, which is a set of parameters in accordance with which the worker’s access to his target is achieved.

    The variables that make up the access model include:
    • Transfer Request Size - the minimum data block that the test application can access.
    • Percent Random / Sequential Distribution - percentage of random queries. The remaining requests, respectively, are sequential.
    • Percent Read / Write Distribution - percentage of read / write requests.

    Another important variable that is not directly included in the access model - # of Outstanding I / Os - determines the number of simultaneous I / O requests for a given worker and, accordingly, disk loading. If this parameter is set to 1, then we get a measurement of random access time. A value of 4 corresponds to loading a completely basic application. On real applications, this parameter takes a value of 30-50. A parameter value greater than 100 corresponds to a very serious disk load, such as during defragmentation. In accordance with this test will be carried out using the five following values ​​of this parameter.
    • Linear - 1 Outstanding I / O
    • Very Light - 4 Outstanding I / Os
    • Light - 16 Outstanding I / Os
    • Moderate - 64 Outstanding I / Os
    • Heavy - 256 Outstanding I / Os

    The article tested three commonly accepted access models.
    File Server - this pattern simulates the operation of a hard disk as a server disk subsystem.
    Workstation - this pattern simulates the work of the hard disk when executing 2D / 3D-design and video editing programs.
    Database - this pattern imitates the work of the hard disk when actively working with databases.

    image
    IOMetr Test Settings

    After running the test, Iometer creates a file with a large number of different numbers: the average number of requests executed per second, the average time to complete the operation, the maximum time to complete the operation, the total number of bytes read and written, and CPU utilization in percent. In order not to clutter up the article, the diagrams will show three numbers for each configuration of the disk subsystem, which are ratings in the File Server, Workstation and Database access models, respectively. The rating is calculated as the arithmetic average of Total I / Os Per Second for all values ​​of the number of simultaneous I / O requests as a percentage of the corresponding result of the hard drive 7200.11 ST31500341AS with a capacity of 1500 GB, which will be compared in the subsequent reviews in the Iometer test.

    Iometer File Server
    After depressing results in PCMark, we will offer the discs to play in the home field for the server Seagate Cheetah field, that is, in the realm of seroser tasks :).

    image
    Results of the File Server script.

    Let me remind you that 100 corresponds to the speed of Seagate Barracuda of the previous generation 7200.11, the new 7200.12, as we see, is slightly faster than it. But this increase in speed fades in comparison with the acceleration obtained from the operation of the SAS-disk. 130% difference for single drives, double the advantage of Cheetah over RAID0 of 7200 rpm drives. If someone lacks this speed, then the second SAS allows you to increase it by another 40%. In general, which of these disks to put in the file server, such a problem simply does not exist - the whole question is how many pieces to take.

    Iometer workstation

    image
    Workstation scenario

    results The results show that it is not worth saving on a disk subsystem in a workstation. 15,000 rpm look very impressive. wheels. Unless RAID0 scalability in this case is worse - only 20%, but this is not so important - regular disks are still far behind. By the way, in this Seagate 7200.12 pattern it turned out to be almost equal in speed to its predecessor, in other tests it usually had a bigger advantage.

    Iometer Database Database

    image
    Script Results Databases

    also like high-speed hard disks - we again see their advantage more than twice. Indeed, they love SAS disks with high loads, they proved to be excellent in all access models in the Iometer test.

    Conclusion
    The conclusion suggests itself that using SAS disks in desktop systems is unjustified. The high price, strict cooling requirements and inadequate performance of many popular test packages make us think that for all the disadvantages the user will not receive noticeable (with the exception of certain specific tasks) performance advantages. So cooler is not always better. And the use by some manufacturers of ready-made SAS-disk systems in their configurations is more like a marketing ploy to attract wealthy customers.

    Source OCClub.ru

    Also popular now: