Server hard disk

The server hard disk, as its name implies, is the hard disk used on the server. If the server is the core of network data, then the server hard disk is the core data warehouse, and all software and user data are stored here. For users, the hard disk data stored on the server is the most valuable, so the reliability of the hard disk is very important. In order to make the hard disk adapt to a large data volume and long work time working environment, the server generally adopts a high-speed, stable, and secure SCSI hard disk.

Today's hard drives are divided into two types: IDE hard disks and SCSI hard disks (There are still some products that support PCMCIA interface, IEEE 1394 interface, SATA interface, USB interface, and FC-AL (Fibre Channel-Arbitrated Loop) Fibre Channel interface. But relatively few); IDE hard disk is our daily use of hard disk, it is not bad because of its low price, so it has been widely used in the PC, most of the hard disk used on personal computers are currently Type of hard disk. Another type of hard disk is a SCSI hard disk (SCSI, Small Computer System Interface, small computer system interface). Due to its good performance, this type of hard disk is commonly used on servers, but at the same time, it is expensive, so it is common. SCSI is not often seen on PCs.

Compared with the hard disk of an ordinary PC, the hard disk used on the server has the following four characteristics.

1. The speed of the hard disk used by the server is fast, which can reach 7200 or 10,000 rpm, or even higher; it also has a large (usually 2MB or 4MB) write-back cache; the average access time is relatively short; Higher transmission rate and internal transfer rate, using standard SCSI hard disks such as Ultra Wide SCSI, Ultra2 Wide SCSI, Ultra160 SCSI, Ultra320 SCSI, etc. The data transfer rate per second can reach 40MB, 80MB, 160MB, 320MB respectively.

2, high reliability because the server hard disk is almost 24 hours of non-stop operation and bear a huge workload. It can be said that if there is a problem with the hard disk, the consequences can be disastrous. Therefore, the current hard drives have adopted SMART technology (self-monitoring, analysis and reporting technology), while hard disk manufacturers have adopted their own unique advanced technology to ensure data security. In order to avoid unexpected losses, server hard drives can generally withstand 300G to 1000G impact.

3. Use of SCSI Interfaces Most of the servers use SCSI hard disks with large data throughput and extremely low CPU occupancy. SCSI hard disks must be used through SCSI interfaces. Some server boards have integrated SCSI interfaces, some have dedicated SCSI interface cards, and one SCSI interface card can accept seven SCSI devices. This is unmatched by the IDE interface.

4. Hot-swappable Hot-swappable (Hot Swap) is a hard disk installation method supported by some servers. It can pull out or insert a hard disk without stopping the server. The operating system automatically recognizes the hard disk changes. This kind of technology is very necessary for a 24-hour server.

When we measure the performance of a server hard disk, we should mainly refer to the following indicators:
Spindle speed Spindle speed is a performance factor that is the most compelling factor in all hard disk drives except for capacity. It is also the first determinant of internal hard disk internal transfer speed and continuous transfer speed. Today, the speed of the hard disk is mostly 5400rpm, 7200rpm, 10000rpm and 15000rpm. From the current situation, the 10000rpm SCSI hard disk has the advantage of cost-effective, is currently the mainstream of hard drives, and 7200rpm and below hard drives gradually fade out of the hard disk market .

The internal transfer rate internal transfer rate is the decisive factor in evaluating the overall performance of a hard disk. Hard disk data transmission rate is divided into internal and external transmission rate; usually external transmission rate is also referred to as burst data transfer rate (Burstdata Transfer Rate) or interface transmission rate, refers to the speed of data output from the hard disk cache, currently using Ultra 160 The external transfer rate of SCSI technology has reached 160MB/s; the internal transfer rate is also called the maximum or minimum sustained transfer rate (Sustained Transfer Rate), which refers to the speed at which the hard disk reads and writes data on the platter. Currently, the majority of mainstream hard disks are at 30MB/inch. s to 60MB/s. Since the hard disk's internal transfer rate is less than the external transfer rate, only the internal transfer rate can be used as a true measure of hard disk performance.

In addition to the single-disk capacity contribution to capacity growth, another important aspect of single-disk capacity is to increase the data transfer speed of hard disks. The increase in the single-disc capacity is due to the increase in the number of tracks and the increase in linear magnetic density in the track. The increase in the number of tracks is greatly advantageous for reducing the seek time of the magnetic head because the radius of the magnetic plate is fixed. An increase in the number of tracks means a reduction in the distance between the tracks, and the magnetic head needs to be transferred from one track to another track. The bit time will be shortened. This will help increase the speed of random data transmission. The increase in the magnetic linear density within the track is directly related to the continuous data transmission speed of the hard disk. The increase in linear density within the track allows more data to be stored in each track, so that more data is read from the head into the hard disk buffer during each circular movement of the disc.

Average seek time The average seek time refers to the time required for the head to move to the track where the data is located. This is an important indicator to measure the mechanical performance of the hard disk. It is generally between 3ms and 13ms. It is recommended that SCSI hard disks with an average seek time greater than 8ms should not be considered. . The average seek time and the average latency (determined entirely by the speed) determine the time at which the hard disk head finds the cluster in which the data is located. This time directly affects the random data transmission speed of the hard disk.

Caching increases the capacity of the hard disk cache is also a shortcut to improve the overall performance of the hard disk. Because the hard disk internal data transmission speed and external transmission speed are different. So you need to cache to do a speed adapter. The size of the cache has a great impact on the sustained data transfer speed of the hard disk. Its capacity is 512KB, 2MB, 4MB, even 8MB or 16MB. For video capture, video editing, etc., which requires a lot of disk input/output work, large hard disk cache is an ideal choice.

Because SCSI has the advantages of low CPU usage, high efficiency of multitasking concurrent operations, many connected devices, long connection distances, etc. For most server applications, SCSI hard disks are recommended, and the latest Ultra160 SCSI controllers are used; for low-end Small server applications can use the latest IDE hard drives and controllers. After determining the interface and type of hard disk, it is necessary to focus on the above-mentioned technical indicators affecting the performance of the hard disk, according to factors such as the speed, single-disc capacity, average seek time, cache, etc., combined with the capital budget, the most appropriate choice of cost performance Hard disk program. In a specific application, the first choice is to use a hard disk with a long life and a low failure rate to reduce the probability and frequency of failures. This involves the hard disk MTBF (mean time between failures) and data protection technology. The larger the MTBF, the better. For example, the hard disk used by Inspur Yingxin Server usually has an MTBF value of more than 1.2 million hours, and SMART (self-monitoring, analysis, and reporting technology) and similar technologies such as seagate and IBM's DST (drive self-test) and DFT (driver) are common to hard disk drives. Health testing) is of great importance for the security of data stored on hard disks.

At present, the main server hard disk manufacturers are mainly seagate (Seagate), IBM and so on.