Raid 0 array of 4 hard drives. Types of RAID and their characteristics
The problem of increasing the reliability of information storage is always on the agenda. This is especially true for large amounts of data, databases on which the operation of complex systems in a wide range of industries depends. This is especially important for high-performance servers.
As you know, the performance of modern processors is constantly growing, which modern processors clearly cannot keep up with in their development.
hard disks. Having one disk, be it SCSI or, even worse, IDE, is already won't be able to decide tasks relevant to our time. You need many disks that will complement each other, replace them if one of them fails, store backup copies, and work efficiently and productively.
However, simply having several hard drives is not enough, you need them integrate into a system, which will work smoothly and will not allow data loss in the event of any disk-related failures.
You need to take care of creating such a system in advance, because, as the famous proverb says, Bye fried the rooster won't bite- they won’t miss it. You may lose your data irrevocably.
This system could become RAID– a virtual storage technology that combines several disks into one logical element. A RAID array is called redundant array independent disks. Typically used to improve performance and reliability.
What is needed to create a raid? At least two hard drives. Depending on the array level, the number of storage devices used varies.
What types of raid arrays are there?
There are basic, combination RAID arrays. The Berkeley Institute in California proposed dividing the raid into specification levels:
- Basic:
- RAID 1 ;
- RAID 2 ;
- RAID 3 ;
- RAID 4 ;
- RAID 5 ;
- RAID 6 .
- Combined:
- RAID 10 ;
- RAID 01 ;
- RAID 50 ;
- RAID 05 ;
- RAID 60 ;
- RAID 06 .
Let's look at the most commonly used ones.
Raid 0
RAID 0 intended to increase speed and recording. It does not increase storage reliability and is therefore not redundant. His other name is stripe (striping - “alternation”). Usually used from 2 to 4 disks. 
The data is divided into blocks, which are written to disks one by one. Speed writing/reading increases by a number of times that is a multiple of the number of disks. From shortcomings One can note the increased likelihood of data loss with such a system. It makes no sense to store databases on such disks, because any serious failure will lead to complete inoperability of the raid, since there are no recovery tools.
Raid 1
RAID 1 provides mirror data storage at the hardware level. Also called an array Mirror, What means « mirror»
. That is, the disk data in this case is duplicated. Can use with the number of storage devices from 2 to 4. 
Speed writing/reading practically does not change, which can be attributed to benefits. The array works if at least one raid disk is in operation, but the system volume is equal to the volume of one disk. In practice, when failure one of the hard drives, you will need to take steps to replace it as quickly as possible.
Raid 2
RAID 2 - uses the so-called Hamming code. Data is split across hard drives similar to RAID 0, and is stored on the remaining drives error correction codes, in case of failure by which you can regenerate information. This method allows on-the-fly find, and then correct system failures. 
Rapidity read/write in this case compared to using one disk rises. The downside is the large number of disks, for which it is rational to use it so that there is no data redundancy, usually this 7 and more.
RAID 3 - in an array, the data is split across all disks except one, which stores the parity bytes. Resistant to system failures. If one of the disks fails. Then its information can be easily “raised” using parity checksum data. 
Compared to RAID 2 no possibility error correction on the fly. This array is different high performance and the ability to use 3 disks or more.
Main minus Such a system can be considered an increased load on the disk that stores parity bytes and low reliability of this disk.
Raid 4
In general, RAID 4 is similar to RAID 3 except difference that parity data is stored in blocks rather than bytes, which allows for increased speed of small data transfers.
Minus The specified array turns out to have a write speed, because write parity is generated on one single disk, just like RAID 3. 
This seems to be a good solution for those servers where files are read more often than written.
Raid 5
RAID 2 to 4 have disadvantages due to the inability to parallelize write operations. RAID 5 eliminates this drawback. Parity blocks are written simultaneously to all disk devices in the array, no asynchrony in the data distribution, which means parity is distributed. 
Number used hard drives from 3. The array is very common due to its versatility And efficiency, the greater the number of disks used, the more economical the disk space will be spent. Speed wherein high due to data parallelization, but performance is reduced compared to RAID 10 due to the large number of operations. If one drive fails, reliability drops to RAID 0. It takes a long time to recover.
Raid 6
RAID 6 technology is similar to RAID 5, but higher reliability by increasing the number of parity disks. 
However, at least 5 disks and a more powerful processor are already required to process the increased number of operations, and the number of disks must be equal to the prime number 5,7,11 and so on.
Raid 10, 50, 60
Next come combinations the previously mentioned raids. For example, RAID 10 is RAID 0 + RAID 1. 
They inherit and advantages arrays of their components in terms of reliability, performance and number of disks, and at the same time efficiency.
Creating a raid array on a home PC
The advantages of creating a raid array at home are not obvious, due to the fact that it uneconomical, data loss is not so critical in comparison with servers, but information can be stored in backup copies, making backups periodically.
For these purposes you will need raid controller, which has its own BIOS and its own settings. In modern motherboards, the raid controller can be integrated to the south bridge of the chipset. But even in such boards, you can connect another controller by connecting to a PCI or PCI-E connector. Examples include devices from Silicon Image and JMicron.
Each controller can have its own configuration utility.
Let's look at creating a raid using the Intel Matrix Storage Manager Option ROM.
Transfer all data from your disks, otherwise during the creation of the array they will be cleared.
Go to BIOSSetup your motherboard and turn on the operating mode RAID for your sata hard drive. 
To launch the utility, restart your PC, click ctrl+i during the procedure POST. In the program window you will see a list of available disks. Click Create Massive. Next select required array level.
In the future, following the intuitive interface, enter array size And confirm its creation.
Today we will talk about RAID arrays. Let's figure out what it is, why we need it, what it is like and how to use all this magnificence in practice.
So, in order: what is RAID array or simply RAID? This abbreviation stands for "Redundant Array of Independent Disks" or "redundant (backup) array of independent disks." To put it simply, RAID array this is a collection of physical disks combined into one logical disk.
Usually it happens the other way around - one physical disk is installed in the system unit, which we split into several logical ones. Here the situation is the opposite - several hard drives are first combined into one, and then the operating system is perceived as one. Those. The OS firmly believes that it physically only has one disk.
RAID arrays There are hardware and software.
Hardware RAID arrays are created before loading the OS using special utilities built into RAID controller- something like a BIOS. As a result of creating such RAID array already at the OS installation stage, the distribution kit “sees” one disk.
Software RAID arrays created by OS tools. Those. During boot, the operating system “understands” that it has several physical disks, and only after the OS starts, through software, the disks are combined into arrays. Naturally, the operating system itself is not located on RAID array, since it is installed before it is created.
"Why is all this needed?" - you ask? The answer is: to increase the speed of reading/writing data and/or increase fault tolerance and security.
"How RAID array can increase speed or secure data?" - to answer this question, consider the main types RAID arrays, how they are formed and what it gives as a result.
RAID-0. Also called "Stripe" or "Tape". Two or more hard drives are combined into one by sequential merging and summing up the volumes. Those. if we take two 500GB disks and create them RAID-0, the operating system will perceive this as one terabyte disk. At the same time, the read/write speed of this array will be twice as high as that of one disk, since, for example, if the database is physically located in this way on two disks, one user can read data from one disk, and another user can write to another disk at the same time. Whereas, if the database is located on one disk, the hard disk itself will perform read/write tasks of different users sequentially. RAID-0 will allow reading/writing in parallel. As a consequence, the more disks in the array RAID-0, the faster the array itself works. The dependence is directly proportional - the speed increases N times, where N is the number of disks in the array.
At the array RAID-0 there is only one drawback that outweighs all the advantages of using it - the complete lack of fault tolerance. If one of the physical disks of the array dies, the entire array dies. There's an old joke about this: "What does the '0' in the title mean? RAID-0? - the amount of information restored after the death of the array!"
RAID-1. Also called "Mirror" or "Mirror". Two or more hard drives are combined into one by parallel merging. Those. if we take two 500GB disks and create them RAID-1, the operating system will perceive this as one 500GB disk. In this case, the read/write speed of this array will be the same as that of one disk, since information is read/written to both disks simultaneously. RAID-1 does not provide a gain in speed, but provides greater fault tolerance, since in the event of the death of one of the hard drives, there is always a complete duplicate of information located on the second drive. It must be remembered that fault tolerance is provided only against the death of one of the array disks. If the data was deleted purposefully, it is deleted from all disks of the array simultaneously!
RAID-5. A more secure option for RAID-0. The volume of the array is calculated using the formula (N - 1) * DiskSize RAID-5 from three 500GB disks, we get an array of 1 terabyte. The essence of the array RAID-5 is that several disks are combined into RAID-0, and the last disk stores the so-called “checksum” - service information intended to restore one of the array disks in the event of its death. Array write speed RAID-5 somewhat lower, since time is spent calculating and writing the checksum to a separate disk, but the reading speed is the same as in RAID-0.
If one of the array disks RAID-5 dies, the read/write speed drops sharply, since all operations are accompanied by additional manipulations. Actually RAID-5 turns into RAID-0 and if recovery is not taken care of in a timely manner RAID array there is a significant risk of losing data completely.
With an array RAID-5 You can use the so-called Spare disk, i.e. spare. During stable operation RAID array This disk is idle and not used. However, in the event of a critical situation, restoration RAID array starts automatically - information from the damaged one is restored to the spare disk using checksums located on a separate disk.
RAID-5 is created from at least three disks and saves from single errors. In case of simultaneous occurrence of different errors on different disks RAID-5 doesn't save.
RAID-6- is an improved version of RAID-5. The essence is the same, only for checksums, not one, but two disks are used, and the checksums are calculated using different algorithms, which significantly increases the fault tolerance of everything RAID array generally. RAID-6 assembled from at least four disks. The formula for calculating the volume of an array looks like (N - 2) * DiskSize, where N is the number of disks in the array, and DiskSize is the size of each disk. Those. while creating RAID-6 from five 500GB disks, we get an array of 1.5 terabytes.
Write speed RAID-6 lower than RAID-5 by about 10-15%, which is due to additional time spent on calculating and writing checksums.
RAID-10- also sometimes called RAID 0+1 or RAID 1+0. It is a symbiosis of RAID-0 and RAID-1. The array is built from at least four disks: on the first RAID-0 channel, on the second RAID-0 to increase read/write speed, and between them in a RAID-1 mirror to increase fault tolerance. Thus, RAID-10 combines the advantages of the first two options - fast and fault-tolerant.
RAID-50- similarly, RAID-10 is a symbiosis of RAID-0 and RAID-5 - in fact, RAID-5 is built, only its constituent elements are not independent hard drives, but RAID-0 arrays. Thus, RAID-50 gives very good read/write speed and contains the stability and reliability of RAID-5.
RAID-60- the same idea: we actually have RAID-6, assembled from several RAID-0 arrays.
There are also other combined arrays RAID 5+1 And RAID 6+1- they look like RAID-50 And RAID-60 the only difference is that the basic elements of the array are not RAID-0 tapes, but RAID-1 mirrors.
How do you understand combined RAID arrays: RAID-10, RAID-50, RAID-60 and options RAID X+1 are direct descendants of the basic array types RAID-0, RAID-1, RAID-5 And RAID-6 and serve only to increase either read/write speed or increase fault tolerance, while carrying the functionality of basic, parent types RAID arrays.
If we move on to practice and talk about the use of certain RAID arrays in life, the logic is quite simple:
RAID-0 We do not use it in its pure form at all;
RAID-1 We use it where read/write speed is not particularly important, but fault tolerance is important - for example, on RAID-1 It’s good to install operating systems. In this case, no one except the OS accesses the disks, the speed of the hard disks themselves is quite sufficient for operation, fault tolerance is ensured;
RAID-5 We install it where speed and fault tolerance are needed, but there is not enough money to buy more hard drives or there is a need to restore arrays in case of damage without stopping work - spare Spare drives will help us here. Common Application RAID-5- data storage;
RAID-6 used where it is simply scary or there is a real threat of death of several disks in the array at once. In practice it is quite rare, mainly among paranoid people;
RAID-10- used where it is necessary to work quickly and reliably. Also the main direction for use RAID-10 are file servers and database servers.
Again, if we simplify it further, we come to the conclusion that where there is no large and voluminous work with files, it is quite enough RAID-1- operating system, AD, TS, mail, proxy, etc. Where serious work with files is required: RAID-5 or RAID-10.
The ideal solution for a database server is a machine with six physical disks, two of which are combined into a mirror RAID-1 and the OS is installed on it, and the remaining four are combined into RAID-10 for fast and reliable data processing.
If, after reading all of the above, you decide to install it on your servers RAID arrays, but don’t know how to do it and where to start - contact us! - we will help you select the necessary equipment, as well as carry out installation work for implementation RAID arrays.
RAID arrays were developed to improve data storage reliability, increase processing speed, and provide the ability to combine multiple disks into one large one. Different types of RAID solve different problems; here we will look at several of the most common configurations of RAID arrays of the same size.
RAID 0
RAID 1
RAID 10
RAID 1E
RAID 5
"Software" implementations of RAID5, built into the south bridges of motherboards, do not have high write speeds, so they are not suitable for all applications.
RAID 5EE
RAID 6
RAID 50
RAID 60
Enthusiasts have an unwritten rule: hard drive Western Digital WD1500 Raptor is the ideal desktop model if you need maximum performance. But not all users can follow this path, since spending $240 on a hard drive with a capacity of only 150 GB is not a very attractive solution. Is the Raptor still the best choice? The price has not changed for many months, and today for that kind of money you can easily buy a pair of 400 GB drives. Isn't it time to compare the performance of modern RAID arrays with Raptor?
Enthusiasts are familiar with Raptor hard drives because it is the only 3.5" desktop hard drive that spins at 10,000 rpm. Most hard drives in this market sector spin at 7,200 rpm. Only high-end hard drives for servers spin faster. WD Raptor hard drives 36 and 74 GB were introduced three years ago. About a year ago it entered the market Western Digital Raptor-X, which provides higher performance, models are also available with a transparent cover that allows you to look inside the hard drive.
After their release, Western Digital Raptor hard drives outperformed all other 3.5" Serial ATA hard drives for desktop PCs, although they were initially positioned for low-cost servers.
A spindle speed of 10,000 rpm offers two significant advantages. Firstly, the data transfer speed increases noticeably. Yes, the maximum sequential read speed is not particularly impressive, but the minimum speed is far superior to any 7,200 RPM hard drive. Additionally, a 10,000 RPM hard drive has less spin-up latency, meaning the drive takes less time to acquire data once the read/write heads are positioned.
The main disadvantage of the WD Raptor is the price - about $240 for the 150 GB model. Among other disadvantages, we note a higher (although not critical) noise level and higher heat generation. However, enthusiasts will easily put up with such shortcomings if this hard drive provides higher performance of the storage subsystem.
If you calculate the cost of storing a gigabyte of data, then Raptor will no longer be so attractive. For $240 you can get a pair of 400 GB hard drives, and the $300 level for a 750 GB Seagate Barracuda 7200.10 is not far away. If you look at the low-end segment, you can get a pair of 160GB 7,200 RPM hard drives for $50 each, which will provide the same capacity as the Raptor, but at more than half the price. Therefore, today even enthusiasts often ask themselves: is it worth taking the WD Raptor, isn’t it better to choose a RAID 0 configuration on two 7,200 rpm hard drives?
RAID 0 does not reduce access time, but it practically doubles the sequential read speed since the data is distributed between two hard drives. The disadvantage is the increased risk of data loss, since if one hard drive fails, the entire array will be lost (however, today there are also options RAID information recovery). Many onboard controllers on high-end motherboards support RAID modes, which are easy to configure and install.
Fast or smart hard drive?
| Performance | Capacity | Data storage security | Price | |
| One hard drive (7,200 rpm) | good | Fair to excellent | Sufficient * | Low to High, $50 to $300 |
| 150GB WD Raptor (10,000 rpm) | Excellent | Sufficient | Sufficient * | High: $240+ |
| 2x 160 GB (7,200 rpm) | Very good to excellent | Good to excellent | Insufficient * | Low to High: $50 per HDD |
| 2x 150 GB WD Raptor (10,000 rpm) | Excellent | good | Insufficient * | High to very high: $240 per drive |
* It should be remembered that any hard drive will fail sooner or later. The technology is based on mechanical components, and their lifespan is limited. Manufacturers indicate the time between failures (MTBF, Mean Time Between Failures) for hard drives. If you have a RAID 0 array installed on two 7,200 rpm hard drives, then the risk of data loss doubles because if one hard drive fails, you will lose the entire RAID 0 array. Therefore, regularly back up important data and create an image of the operating system.
Today you can buy 40-80 GB hard drives for almost pennies, and if you do not have special capacity requirements, then this volume will be enough even today. However, we recommend purchasing hard drives priced between $50 and $70, as you can easily get models with capacities ranging from 120 to 200 GB. Models with 250 and 320 GB have already begun to appear in online stores at a price of less than $100. For the money you spend on a 10,000 RPM WD Raptor, you can easily get 800 GB to 1 TB of capacity on 7,200 RPM hard drives.
If you don't need such high capacity, you can settle for entry-level 7,200 RPM hard drives. Two Western Digital WD1600AAJS drives cost $55 each, and you'll easily get 320GB of capacity in a RAID 0 array. You'll spend half the money and get twice the capacity. How justified are such savings? Let's figure it out.
7,200 or 10,000 rpm? RAID 0 or Raptor?
We decided to test different hard drive configurations. Our testing involves a single WD Raptor WD1500ADFD, a single WD4000KD, a Raptor in RAID 0, and a WD4000 in RAID 0. We decided to use 400GB WD 7,200 RPM hard drives since two of these drives are roughly the same price as one Raptor . Let's see how well the "budget" RAID array performs compared to a single Raptor.
The WD4000KD is equipped with 16 MB of cache and has a Serial ATA/150 interface. The main difference compared to the 10,000 rpm WD Raptor is performance and capacity. The Raptor comes in at a significant cost per gigabyte of storage, which is at least six times higher than the 400GB WD4000KD. Tests will show how big the performance differences are. At the time of publication, the price of the WD4000KD Caviar was $130.
The Raptor is the undisputed performance champion in the desktop market, but it is also the most expensive hard drive. The WD1500 Raptor uses a Serial ATA/150 interface, which is still quite sufficient. If you look at the test results, no other hard drive can beat the Raptor, even with a SATA 300 MB/s interface. In general, SATA interface speed should not factor into your purchasing decision. At the time of publication, the price of the WD1500ADFD Raptor was $240.
This configuration should take on the WD1500 Raptor. Can two WD4000KD hard drives in a RAID 0 array beat the Raptor?
This scenario is the most expensive in our testing because it requires two WD Raptor hard drives, but it is still very interesting. Two 10,000 rpm Raptor hard drives in a RAID 0 array should literally destroy everyone.
RAID 0
Performance
In theory, RAID 0 is ideal for increasing performance because the sequential data transfer rate scales almost linearly with the number of hard drives in the array. Files are distributed block-by-block across all hard drives, that is, the RAID controller writes data almost simultaneously to several hard drives. RAID 0 data transfer speeds increase noticeably in almost all scenarios, although access times do not decrease. In real-world tests, access times in RAID 0 arrays even increase, albeit very slightly, by about half a millisecond.
If you build a RAID configuration on several hard drives, the drive controller may become the bottleneck. A regular PCI bus can transfer a maximum of 133 MB/s, which is easily absorbed by two modern hard drives. Serial ATA controllers that are included in the chipset generally provide higher throughput, so they do not limit the performance of RAID arrays.
We got up to 350 MB/s on four WD Raptor hard drives with 10,000 rpm on chipsets with Intel ICH7 and ICH8 southbridges. An excellent result that is very close to the total throughput of four separate hard drives. At the same time, the nVidia nForce 680 chipset showed a maximum of 110 MB/s, alas. It seems that not every integrated RAID controller is capable of providing high performance RAID arrays, even if technically it is possible.
Comparison of RAID modes
It should be noted that RAID 0 doesn't really cover the idea of RAID arrays, which stands for Redundant Arrays of Independent/Inexpensive Drives. Redundancy means storing data in at least two places so that it survives even if one hard drive fails. This is what happens, for example, in the case of a RAID 1 array, in which all data is mirrored on a second hard drive. If one of the hard drives “dies,” you will only know about it from the RAID controller messages. RAID 5 is much more complex and is aimed at the professional sector. It works like a RAID 0 array, distributing data across all hard drives, but with redundancy information added to the data. Therefore, the net capacity of a RAID 5 array is equal to the total capacity of all hard drives except one. Redundancy information is not written to one hard drive (as in the case of RAID 3), but is distributed across all drives so as not to create a bottleneck when reading or writing redundancy information to one HDD. A RAID 5 array, quite understandably, requires at least three hard drives.
Risks and side effects
The main danger for a RAID 0 array is the failure of any hard drive, since the entire array is lost. That is why the more disks in a RAID 0 array, the higher the risk of losing information. If three hard drives are used, then the likelihood of losing information is three times greater than with one drive. This is why RAID 0 is not a good option for users who need a reliable system and cannot afford a single minute of downtime.
Even if you buy a powerful and expensive separate RAID controller, you will still be dependent on the hardware. Two different controllers may support RAID 5, but the specific implementation may be very different.
Intel Matrix RAID: You can create different RAID arrays on the same set of hard drives.
If the RAID controller is smart enough, it may allow two or more RAID arrays to be installed on one set of hard drives. Although each RAID controller can support multiple RAID arrays, this most often requires different sets of hard drives. Therefore, the Intel ICH7-R and ICH8-R southbridges turned out to be very interesting: they support the Intel Matrix RAID function.
A typical implementation would be two RAID arrays on two hard drives. The first third of the capacity of the two hard drives can be allocated to a fast RAID 0 array for the operating system, and the rest to a RAID 1 array for storing important data. If one of the hard drives fails, the operating system will be lost, but important data that is mirrored to the second hard drive will be preserved thanks to RAID 1. By the way, after installing Windows, you can create an image of the operating system and store it on a reliable RAID 1 array. Then, if the hard drive fails, the OS can be quickly restored.
Please be aware that many RAID arrays require a RAID driver (such as Intel Matrix Storage Manager) to be installed, which can create problems during system boot and recovery. Any boot disk you use for recovery will need RAID drivers. Therefore, save the driver floppy disk for such a case.
Test configuration
Configuration for Low Level Tests
| Processors | 2x Intel Xeon (Nocona core), 3.6 GHz, FSB800, 1 MB L2 cache |
| Platform | Asus NCL-DS (Socket 604), Intel E7520 chipset, BIOS 1005 |
| Memory | Corsair CM72DD512AR-400 (DDR2-400 ECC, reg.), 2x 512 MB, latencies CL3-3-3-10 |
| System hard drive | Western Digital Caviar WD1200JB, 120 GB, 7200 rpm, 8 MB cache, UltraATA/100 |
| Drive controllers | Intel 82801EB UltraATA/100 Controller (ICH5) Silicon Image Sil3124, PCI-X |
| Net | Built-in Broadcom BCM5721 Gigabit Ethernet controller |
| Video card | Built-in ATi RageXL, 8 MB |
| Tests and settings | |
| Performance tests | c"t h2benchw 3.6 PCMark05 V1.01 |
| I/O tests | IOMeter 2003.05.10 Fileserver-Benchmark Webserver-Benchmark Database-Benchmark Workstation-Benchmark |
| System software | |
| OS | Microsoft Windows Server 2003 Enterprise Edition, Service Pack 1 |
| Platform Driver | Intel Chipset Installation Utility 7.0.0.1025 |
| Graphics driver | Default Windows Graphics Driver |
Configuration for SYSmark2004 SE
| System hardware | |
| CPU | Intel Core 2 Extreme X6800 (Conroe 65 nm, 2.93 GHz, 4 MB L2 cache) |
| Motherboard | Gigabyte GA-965P-DQ6 2.0, chipset: Intel 965P, BIOS: F9 |
| General hardware | |
| Memory | 2x 1024 MB DDR2-1111 (CL 4.0-4-4-12), Corsair CM2X1024-8888C4D XMS6403v1.1 |
| Video card | HIS X1900XTX IceQ3, GPU: ATi Radeon X1900 XTX (650 MHz), memory: 512 MB GDDR3 (1550 MHz) |
| Hard drive I | 150 GB, 10,000 rpm, 8 MB cache, SATA/150, Western Digital WD1500ADFD |
| Hard drive II | 400 GB, 7,200 rpm, 16 MB cache, SATA/300, Western Digital WD4000KD |
| DVD-ROM | Gigabyte GO-D1600C (16x) |
| Software | |
| ATi Drivers | Catalyst Suite 7.1 |
| Intel Chipset Drivers | Software Installation Utility 8.1.1.1010 |
| Intel RAID Drivers | Matrix Storage Manager 6.2.1.1002 |
| DirectX | 9.0c (4.09.0000.0904) |
| OS | Windows XP, Build 2600 SP2 |
| Tests and settings | |
| SYSmark | Version 2004 Second Edition, Official Run |
Well, we'll have to move on to the results of the battle between the current 150 GB WD Raptor hard drives and 400 GB WD4000KD hard drives in a RAID 0 array. The result was surprising. While the WD Raptor remains by far the fastest desktop Serial ATA hard drive, RAID 0 comes out on top in most benchmarks outside of access time and I/O performance. The cost of storing a gigabyte of data on the Raptor is most questionable, since you can buy three times the capacity 7,200 rpm hard drive for half the price. That is, at the price of a gigabyte, Raptor today loses six times. However, if you're concerned about data security, think twice about choosing a RAID 0 array of two cheap 7,200 RPM hard drives over the WD Raptor.
In the coming months, the price of 500 GB hard drives will drop below $100. But the requirements for available space to store high-definition videos, music and photos will increase. Finally, the recording density of hard drive platters continues to increase, so higher-performance 7,200 rpm models will soon be available. In the future, the attractiveness of the Raptor will decrease.
It seems to us that Western Digital should change the pricing policy of the Raptor lineup, since the performance gains come at the expense of large compromises in hard drive capacity. And, I must say, such compromises will not seem justified to everyone. We'd like to see an updated 300GB Raptor hard drive, which could also double as a hybrid hard drive with built-in flash memory for Windows Vista.
When creating a file server or a productive workstation, you often have to face the problem of choosing a disk subsystem configuration. Modern motherboards, even at the budget level, offer the ability to create RAID arrays of all popular levels; we should also not forget about the software implementation of RAID. Which option will be more reliable and productive? We decided to conduct our own testing.
Test bench
As a rule, in small and medium-sized businesses for the role of file servers, department-level servers, etc. a regular PC is used, assembled from ordinary, budget components. The purpose of our testing was to study the performance of the disk subsystem assembled using the chipset's RAID controller and compare it with software implementations of RAID arrays (OS tools). The reason for testing was the lack of widely available objective tests of budget RAIDs, as well as a large number of “myths and legends” on this issue. We did not specifically select iron, but used what was at hand. And several ordinary PCs were at hand for the next implementation, one of which was used as a test bench.
PC configuration:
- Motherboard: ASUS M4N68T-M SocketAM3
- Processor: CPU AMD ATHLON II X2 245 (ADX245O) 2.9 GHz/ 2MB/ 4000MHz Socket AM3
- RAM: 2 x Kingston ValueRAM
DDR-III DIMM 1Gb - Hard drives: HDD 320 Gb SATA-II 300 Western Digital Caviar Blue
7200rpm 16Mb - Operating system: Windows Server 2008 SP2 (32-bit)
- File system: NTFS
The disk subsystem was configured as follows: the operating system was installed on one disk, and a RAID array was assembled from two or three others.
Testing methodology
We chose Intel NAS Performance Toolkit as test software; this package represents a set of tests that allows us to evaluate the performance of the disk subsystem on basic typical tasks. Each test was performed five times and the final result represents the average. We took the performance of a single hard drive as a standard.
We tested RAID0, RAID1 and RAID5 arrays, and RAID5 was tested both in normal mode and in emergency mode, with one disk removed. Why did we only test this array in emergency mode? The answer is simple: for RAID0 such a mode does not exist; if any of the disks fails, the array is destroyed, and the only remaining RAID1 disk will be no different from a single disk.
We tested both hardware and software implementations; initially we also measured the average CPU load, since there is an opinion that software RAID heavily loads the processor. However, we refused to include this measurement in the test results; the load on the processor turned out to be approximately equal and amounted to about 37-40% for a single disk, RAID0, RAID1 and 40-45% for RAID5.
File operations
The classic operations for any drive are read and write operations. Intel NASPT evaluates these parameters in four tests: copying a 247 MB file and 44 folders containing 2833 files with a total volume of 1.2 GB to the drive and back.
Read/Write Files
If we pay attention to the results of the reference disk, we will see that the write speed is almost twice (89%) higher than the read speed. This is due to the peculiarities of the file system and this fact should also be taken into account. RAID0 (striped array), regardless of implementation method, showed 70% higher performance than a single disk, while the speed parameters of RAID1 (mirror) are completely identical to it.
RAID5 deserves a special mention; the write speed to it is unacceptably low, the slowdown is up to 70%, while the read speed is not inferior to the fast RAID0. This may be due to a lack of computing resources and imperfect algorithms, because when recording, additional resources are spent to calculate the checksum. If one of the disks fails, the recording speed drops; the decline in the hardware solution is less pronounced (15%) than in the software solution (40%). In this case, the reading speed drops significantly and corresponds to the speed of a single disk.
Read/write folders
Anyone who has tried to copy a scattering of small files knows that it is better to pre-pack them into an archive, it will be much faster. Our tests only confirm this rule of thumb: reading a scattering of small files and folders is almost 60% slower, reading a large file, writing speed is also slightly (10%) lower.
RAID0 gives a much smaller advantage on write operations (30-40%), and on read operations the difference can be completely neglected. RAID1, as expected, does not bring us any surprises, going head to head with a single disk.
RAID5 shows much better results on small files, but still remains inferior to a single disk by an average of 35%. The read speed is no different from other configurations; we are inclined to believe that in this case the limiting factor is the random access time of the hard drive. But when we removed one disk from the array, we got a very unexpected result, which forced us to double-check it several times, including on another model of hard drive (500 Gb Seagate/Maxtor Barracuda 7200.12/DiamondMax 23<3500418AS>7200rpm 16Mb). The fact is that the write speed of the hardware array dropped sharply (almost three times), and the write speed of software RAID5, on the contrary, increased, perhaps this is due to the software implementation algorithm of the array. And yet we prefer to leave this “phenomenon” without comment.
Working with applications
The following tests reflect the performance of the disk subsystem when working with various applications, primarily office ones. The first test (Content Creation) reflects the use of the disk for storing and working with data; the user creates, opens, and saves documents without showing much activity. The most powerful test is Office Productivity, it simulates active work with documents, searching for information on the Internet (the browser cache is reset to the drive), with a total of 616 files in 45 directories with a volume of 572 MB. The last test - working with a photo album (mainly viewing), is more typical for home use, includes 1.2 GB of photos (169 files, 11 directories).
Work with documents
When working with single files, RAID0 is quite predictably almost twice as fast as RAID1 and a single hard drive (Content Creation test), but during active work it loses all its advantages; in the Office Productivity test, RAID0, RAID1 and a single hard drive show the same results.
RAID5 is a clear outsider in these tests; on single files, the performance of the array is extremely low, and the hardware implementation shows a much better (but still extremely low) result. During active office work, the results are much better, but still lower than those of a single disk and simpler arrays.
Working with photos
In this mode, all arrays showed approximately the same result, comparable to the performance of a single disk. Although RAID5 showed a slightly lower result, although in this case the lag can hardly be noticed with the naked eye.
Multimedia
And finally, multimedia tests, which we divided into two parts: playback and recording. In the first case, HD video is played from the drive in one, two and four streams simultaneously. In the second, recording and simultaneous recording and playback of two files is performed. This test is applicable not only to video, as it characterizes the general processes of linear writing/reading from a disk array.
Playback
RAID0
This type of disk array is a confident leader when working with large files and multimedia. In most cases, it allows you to achieve a significant advantage (about 70%) compared to a single disk, but it has one significant drawback - extremely low fault tolerance. If one disk fails, the entire array is destroyed. It has no special advantages when working with office applications and photos.
Where can RAID0 be used? First of all, on workstations that, due to the nature of their tasks, have to work with large files, for example, video editing. If fault tolerance is required, you can use RAID10 or RAID0+1, which represent a striped array of two mirrors or a mirror of striped arrays, these RAID levels combine the speed parameters of RAID0 and the reliability of RAID1, the disadvantages include significant overhead costs - only half the capacity of the incoming disks is used for storage to an array.
RAID1
The “mirror” does not have any speed advantages over a single disk; the main task of this array is to ensure fault tolerance. Recommended for use when working with office files and small files, i.e. for those tasks where the difference between faster arrays is not so great. Not bad for working with 1C:Enterprise 7.7 in file mode, which, by the nature of working with the disk, is a cross between Office Productivity and Dir copy from / to NAS. For more productive tasks it is not recommended; here it is worth paying attention to RAID10 and RAID0+1.
RAID5
We would not recommend using this type of array in budget systems; in write operations, RAID5 is significantly inferior to even a single hard drive. The only area where its use will be justified is the creation of media servers for storing multimedia data, the main mode of which is reading. Here, such parameters as high read speed (at the RAID0 level) and lower overhead costs for ensuring fault tolerance (1/3 of the array capacity) come to the fore, which gives a good benefit when creating storage of significant volumes. However, it should be remembered that attempting to write to the array leads to a sharp decrease in performance, so uploading new data to such media servers should be done during off-peak hours.
Hardware or software?
The test results did not reveal any noticeable advantages or disadvantages for both implementation options, except for RAID5, the hardware version of which showed better results in a number of cases. Therefore, other features should be taken into account. Such as compatibility and portability.
Hardware RAID is implemented by the south bridge of the chipset (or a separate controller) and requires support from the OS or loading of drivers at the installation stage. The same fact often makes it impossible to use a number of disk and system utilities that use their own boot disks; if their boot loader does not support a RAID controller, then the software simply will not see your array.
The second drawback is that you are tied to a specific manufacturer; if you decide to change the platform or choose a motherboard with a different chipset, you will have to copy your data to an external drive (which in itself can be problematic) and reassemble the array. The main trouble is that if the motherboard unexpectedly fails, you will have to look for a similar model to gain access to your data.
Software RAID is supported at the OS level, so it is largely devoid of these disadvantages, the array is easy to assemble and easily transferred between hardware platforms, in the event of hardware failure, access to data can be easily obtained on another PC that has a compatible version of Windows (younger editions do not support dynamic disks).
Among the disadvantages, it should be noted that it is impossible to install Windows on RAID0 and RAID5 volumes, for the reason that installing Windows on a dynamic volume is only possible when this volume has been converted from a basic boot or system volume. You can read more about dynamic volumes.