Database server installation Deployment in Mycollege.net

Server | data | Database server installation deployment in database Mycollege.net

Directory



A preface to ... ..................... ...... ..... .......... ....... ....... ....... ....... ... 1



Two. Apply the meaning of ... ...... ...... ...... ..... ..... ..... ............. ....... ....... ....... ... 1



Three. Related technology ... ....... ....... ...... ...... ...... ... ............. ....... ....... ....... ... 2



Four. System configuration ... ..... ....... ..... ..... ...... ...... ... ............. ....... ....... ....... ... 6



Five. The characteristics of the system ... ..... ..... ..... ..... ..... ...... ... ............. ....... ....... ....... ... 8



Six. A summary of ... ....... ...... ...... ...... ........ ... .......... ....... ....... ....... ....... ... 9























































Summary

The system uses the SAN storage technology to the database server, improves the system's overall speed by increasing the speed of the visit, and provides a wide space for the system to expand.

This is system, which use SAN storage technic,used the system's database server to improve speed. IT provides the system a wide expand area.

Keywords SAN, disk array, RAID

I. Preface

This system includes the user unified authentication system and the public resource management system, in which the authentication system mainly stores the user login username and the password information, mainly relates to the database query operation, does not have one information to be accessed by multiple users, The speed of data access is not very high only when multiple users are logged in at the same time. The public resource management system stores student numbers, name, elective information, including some of the teacher's relevant information, in addition to the very likely multi-user access to the entire system, there may be multiple users of one of the students or teachers access to information, access to storage equipment, speed, The response time requirement is quite high. And the data of these two systems affect the immediate interests of students and teachers, accidental data loss or damage will have an unpredictable impact on the interests of users, and even lead to the chaos of school management and other serious consequences, so the security of data and its importance. At the same time, because of the increase in registration information, such as new entrants, and so on, there is a higher demand for storage capacity, so the expansion of the system is a worthwhile consideration, so we take a high read-write speed, high scalability of the SAN storage system.

Two. Application significance

SAN storage System plays an important role in improving the performance of the whole system in computer systems with large data storage, fast reading and writing speed and high real-time requirements. The system uses the disk array through the controller, as a local hard disk, and then in these "local hard disk" sharing, to read the array of data server (workstation) to the disk diffraction to the local hard disk, when reading data, The data from the disk array is transmitted directly from the fibre switch to the appropriate server (workstation). Due to the use of optical fiber direct transmission, fast, and each server (workstation) is a dedicated optical fiber transmission, not because of other workstations or the server's reading and writing, the workstation (server) on the disk array data read and write, real-time high, transmission speed is also stable. The process of reading data is done by the disk array controller, which greatly reduces the burden on the server and allows the server to do other more important work.

Three. Related Technologies

The technology used is primarily the disk array principle:

Technology for disk arrays: A disk array is an array of multiple disks, used as a single disk, which stores data in a segmented way on different disks, and when it accesses data, the associated disks in the array act together to significantly reduce data access time and better space utilization. The different technologies used by the disk array, called RAID level, different level for different systems and applications to solve the problem of data security.

Generally high-performance disk arrays are achieved in the form of hardware, further to the disk cache control and disk array combination of a controller or control card, for different users to solve the disk output into the system of four major requirements:
(1) Increase the access speed;

(2) fault tolerance, that is, security;

(3) Effective use of disk space;

(4) Try to balance the CPU, memory and disk performance differences, improve the overall performance of the computer.

The principle of the disk array:

Different technologies used in the disk array for different applications, known as RAID level, RAID is the abbreviation for redundant Array of inexpensive disks, and each level represents a technique that is currently used for RAID 0~raid 5. This level does not represent the high and low levels, levels 5 is not higher than levels 3,level 1 is not lower than level 4, as for the choice of a RAID level products, pure depending on the user's operating environment and application, and levels are not necessarily the relationship. RAID 0 and RAID 1 are available for PCs and PC-related systems such as small network servers and workstations that require high disk capacity and fast disk access, because they are cheaper, but because the average person is not aware of the disk array, and do not see the disk array on their value, the market has not yet opened; Raid 2 and RAID 3 are suitable for large computer and image, CAD/CAM processing; RAID 5 is more for OLTP, because of the urgent need of financial institutions and large data processing centers, the use of more and more famous, but also formed a lot of people to the disk array misunderstanding, the disk array is not to RAID 5; RAID 4 is less used because they have something in common, while raid 4 has its innate limitations. Before introducing each RAID level, look at the two basic technologies that form the disk array:

As shown in the following illustration, the Dftraid disk array controller joins four disks:




These four disks form an array, the controller of the disk array is to treat the four disks as a single disk, such as the C: disk in the DOS environment, to extend the small disk to a large capacity single disk, the user does not have to plan the distribution of data on each disk, and increase the utilization of disk space. Dftraid SCSI disk arrays are more than dozens of disks, forming arrays of up to t or even dozens of T, allowing for almost unlimited disk capacity, while each disk is faster than a single disk for taking and saving.

The disk array treats multiple disks of the same array as a single virtual disk, so their data is stored sequentially in the disk array in a segmented fashion, as shown in the following figure

Disk 0

Disk 1

Disk 2

Disk 3

A0-a1

B0-b1

C0-c1

D0-d1



A2-a3

B2-b3

C2-c3

D2-d3



A4-a5

B4-b5

C4-c5

D4-c5



A6-a7

B6-b7

C6-c7

D6-d7




The data is segmented as needed, starting with the first disk, putting it on the last disk, and then going back to the first disk until the data is distributed.

The following are mainly about several types of raid:

RAID0:

RAID 0 has the highest efficiency in the output of the disk. Disk arrays are more efficient because, in addition to data segmentation, it can perform multiple output requirements at the same time, because each disk in the array can be operated independently, segmented on different disks, different disks can be read and write simultaneously, and can be accessed concurrently in cache memory and disk. However, only the hardware's disk array can perform this performance.

RAID 1

RAID 1 is a technique for using disk mirroring. Its way is to add an extra backup disk to the working disk, two disks are stored in exactly the same data, and the data is written to the working disk while it is written to the backup disk. Unlike normal disk mirroring, RAID 1 has no working and backup disks, multiple disks can act simultaneously with overlapping reads, and even different mirrored disks can be written at the same time, which is an optimal way to call load balancing. For example, there are multiple users at the same time to read data, the system can drive each other mirrored disks, while reading data to reduce the system load, increase I/O performance.

RAID 2

RAID 2 is the separation of data into bits or blocks, the addition of the sea plaintext, the disk array in the interval to write to each disk, and the same address, that is, on each disk, its data are in the same track and sector. The design of RAID 2 is a technology that uses coaxial synchronization, when accessing data, the entire disk array moves together, parallel access in the same position of each disk, so there is the best access time, the bus is a special design, with large bandwidth in parallel transmission of data access, so have the best transmission time. RAID 2 has the best performance in large file access applications, however, if the file is too small, its performance will be pulled down, because the disk access is in sector units, and RAID 2 access is the parallel action of all disks, but also for the unit element of access, so less than a sector of the amount of data will be greatly reduced performance.

RAID 3

RAID 3 data is stored and accessed in the same way as RAID 2, but in terms of security it replaces the sea plaintext with error correction and detection, so only an additional school check disk is required. The calculation of parity value is the logical operation of XOR on the relative position of each disk, the result is then written to the parity disk, and any modifications to the data are calculated with parity, such as a disk failure, and the entire disk array (including parity disk) needs to be recalculated once the new disk is replaced. Recover the data from the failed disk and write to the new disk, such as a parity disk failure, recalculate the parity value to achieve fault tolerance.

RAID 4

RAID 4 is segmented by sector, and partitions of each disk at the same location form a checksum disk fragment, which is placed on the checksum disk. This way can be performed on different disks parallel to different read lives today, significantly improve the read performance of the disk array, but when writing data, due to the verification disk, only one time, start all the disk read data to form the same checksum segmentation of all data segments, and the data to be written to verify the calculation and write. Even so, the writing of small files is still faster than RAID 3, because the checksum computation is simpler rather than bit level, but the checksum disk is a bottleneck in RAID 4, which degrades performance.

RAID 5

RAID5 avoids the bottleneck of RAID 4, the method is to not check the disk and officers transferred Guevara the data is placed in a circular manner on each disk, the first partition of the disk array is a checksum value, the second disk to the next disk is the partition of the first disk is the data, then the second disk segment is the checksum value, To retrace the second disk from the third disk is the number of segments. According to, and so on, until the end. The checksum value is computed by the segmented data in the same location on each disk. This method can greatly increase the access performance of small files, not only can be read at the same time, it is possible to execute multiple write actions simultaneously.

The above strategies are compared: RAID0 disk utilization, data read and write speed, but the security of data is poor, RAID1 data read-write speed, data security has a good guarantee, the disadvantage is that the low utilization of disk RAID2 to RAID5 can be said to be a combination of RAID0 and RAID1, read-write speed, data security also has a good guarantee, the disk utilization ratio is high, but RAID2 and RAID3 suitable for mass data in bulk reading and writing, in a large number of small data read and write performance is not very good RAID4 solves the above problem, but the check disk is easy to form the bottleneck of communication, and RAID5 's appearance solves this problem. Therefore, the information stored in the user authentication system and the public resource management system is a large number of small pieces of data, and the data security requirements are high, so it is not suitable to use RAID0, RAID2 and RAID3 read, the system, especially the public resource management system, It is possible to have multiple users at the same time to a student or teacher access to information, if the use of RAID1 read and write, when this situation can be mirrored to each other to read the disk, compared to RAID5 will undoubtedly greatly improve the performance of the system. Therefore, the RAID1 strategy is adopted in the read-write strategy of the disk array. (See also http://www.dft.com.cn for more technology)

In order to ensure the uninterrupted operation of the system, the redundancy of system equipment is used to guarantee it. The disk array uses the dual controller, in the usual case, the dual controller works together, enhances the input output data the speed, enhances the system the performance, simultaneously can reduce the system service interruption because of the controller fault greatly, similarly, the server also makes the way of the Cluster service.

Four. System Configuration

Wiring diagram of the system (Fig. A)

All kinds of servers in addition to the basic configuration, here only describes its special hardware configuration. Database server: Ethernet card (100M), Fibre Channel card (FC card). Web server: Ethernet card (100M), Fibre Channel card, Fibre optic card (additional Ethernet card is required if the Cluster service is to be made). Ethernet Switches and FC switches (for example, 16-port switches) with Cisco or Vixel switches. Ethernet switches do not require any other special configuration, if you want to use the same switch to connect other networks that are not part of the system, it is a good idea to isolate the server connected to the storage network to the switch into a single VLAN to avoid interference. For FC switches, can be configured with software such as San Insite, assuming that the optical fiber connecting the disk array controller is connected to the 1th and 2nd interface of the FC switch, the interface of the fibre connecting the other server should be set up to communicate with the 1th and 2nd interfaces and cannot communicate with each other. This avoids interference between the servers. When configuring a controller, be careful to shield the LUN from servers that do not have to read the LUN to improve the security of the data.

Software configuration: Server Unified installation of Windows ADVANCE Server, installed operating system and hardware drivers, install Sanergy, restart the server, and then test the server on the disk array of read data and write data speed, If the system is working for around 75M, install SQL Server for the database server, and import the data. For Web servers, install the IIS service. Ethernet card is the storage network of Ethernet switches, fiber card access to the local area network or wide area network, responsible for communication with the user, so Ethernet card IP and Fiber optic network adapter settings must be set to not the same network segment. For a server such as public resource management, multiple servers can be clustered into a cluster service to improve performance and stability.

Five. Characteristics of the system

Disks on a disk array can be divided into different volumes, in order to ensure data security, the user identity authentication system data (user name and password) placed on the top of a LUN, while the public resource management system related to the data on the other LUNs, the server does not pass access to the relevant LUNs, This allows the destruction of data that can be caused by preventing illegal access to different servers. As a result, the disk array is divided into at least two LUNs (LUN0 and LUN1), storing the relevant data for each of the two subsystems, and dividing the other LUNs to store other data if the space is large enough. These volumes are divided into disks on the server and set to be shared, and then other Web servers are easy to use and intuitive to access disks on the disk array by diffraction them to local disks. The transmission of data is through specialized optical fiber transmission, rather than through Ethernet (this feature is implemented by Sanergy, both the database server and the Wed server need to install sanergy), fast, and there will be no contention for the transmission channel, so the response time is short, the data read and write speed is also stable. Disk-array disks support hot-swappable technology, in the event of disk corruption, you can unplug the damaged disk without system downtime, replace the new disk, rebuild the data on the damaged disk, and other devices, such as the disk array, the server, and other power supplies also support hot-swappable technology, It can ensure the continuity of the system's work and the security of the data when the system is running and the damaged parts are replaced. As the volume of data grows, when a single disk array does not meet the need to expand the new disk array to increase storage capacity, and without adding new expensive controller, so that the user provides a lot of space for expansion, while the cost of expansion is greatly reduced.

Six. Summary

This graduation design is mainly used during the internship to learn the network storage technology, in the graduation design to carry out a practical application, is to study at the time of the assessment. In this design query in various aspects of the process so that I have a more profound understanding of networked storage technology, mastered a lot of data security and access to the speed of the idea. At the same time, through the cooperation with other students of the group, personally realize the main nature of collaborative cooperation. In the completion of this article, heartfelt thanks to the guidance of the teacher's meticulous guidance and the group of the students to my strong help, this design can be successfully completed, Chengdu Cable-PUI Digital Technology Co., Ltd. for me to provide internship opportunities have an indispensable role in this express heartfelt thanks.




(Figure A)