Source: www.918x.com
In the face of the current DDoS and hacker data, the cluster system will greatly improve the security of the server.
In the process of invading the virtual host, we often have some incredible things happen, as clearly written to the server, but in the web is not found.
Our high-performance virtual host is actually not a server, but a cluster. The interconnection of the host in the cluster is not as simple as the LAN, and they achieve data synchronization, memory synchronization, etc. through the cluster technology Cluster and the cluster software.
1. Basic concept of the cluster
Cluster cluster technology can be defined as follows: a set of independent servers appear as a single system in the network and managed in a single system mode. This single system provides a high reliability service for the customer workstation.
In most modes, all computers in the cluster have a common name, and services running on any system in the cluster can be used by all network customers. Cluster must be able to coordinate errors and failures of each separated component and can be transparent to add components to the Cluster.
A Cluster contains multiple (at least two) servers with shared data storage space. When any server runs an app, the application data is stored in the shared data space. The operating system and application files for each server are stored on their respective local storage spaces.
Interior Node Server communicates with each other in an internal local area network. When a node server fails, the running application on this server will be automatically taken over another node server. When an application service fails, the application service will be restarted or take over another server. When any of the above failures occurs, the customer will be able to connect to the new application service very quickly.
2, two common modes of cluster
First type: mirror server dual machine
The mirror server dual-machine system in the cluster is the simplest and lowest price of hardware configuration. It usually requires two servers in the hardware configuration of the mirroring service. There are independent operating system hard drives and data storage hard drives, each server has Network cards connected to the client, another network card that matches the mirror card or the mirror function.
The mirror server has a simple configuration, convenient use, and low price, but because the mirror server needs to use the network mode image data, the data synchronization is realized by the mirroring software, so it is necessary to occupy the CPU and memory resources of the network server, the performance of the mirror server is longer than single The server has a low performance.
Some mirror server cluster systems use memory mirroring technology, the advantage of this technology is that all applications and network operating systems are synchronized on both servers. When the host fails, the backup machine can take over in the case of almost no feeling. All applications. Because the memory of the two servers is exactly the same, when the system application has defects, the two servers will be simultaneously downtime. This is also a memory mirror card or network card implementation data synchronization. Two servers in the big data amount read and write will generate data asynchronous in some states, so the mirror server is suitable for those who have less budget and is not high for cluster systems. .
Hardware configuration example:
• Two network servers
• Server operating system hard disk two pieces
• Server data storage hard drives need to be determined by users
• Server mirror card (some software can use standard network card) two pieces
• Two blocks of network service network card
Second type two-machine and disk array cabinet
Compared with the mirror server dual-machine system, the interconnection structure of the dual-machine and disk container has more than a third-party disk array cabinet. At present, many companies such as Hairi, Exceptions have produced disk array cabinets, in disk arrays. A disk array control card is installed in the cabinet, and the array cabinet can directly configure the hard disk in the cabinet into a logical disk. Disk array cabinets are connected to the server with a normal SCSI card on the server via the SCSI cable, and the system administrator needs to configure disk arrays directly on disk cabinets.
Double-machine and disk array cabinet interconnection structure does not adopt a memory mirror technology, there is a need to have a certain switching time (usually 60-180 seconds), which can have all downtimes for the system due to the application itself, Since all data is stored in the middle disk array cabinet, when the workmaker fails, the backup machine is replaced, and the data is read from the disk array, so it does not generate the problem of data disagreement, because of this The solution does not require a network image synchronization, so the performance of this cluster program server is much higher than the mirror server structure. The disadvantage of the interconnection of the dual-machine and disk array cabinet is that there is a single-point fault defect in the system, and the so-called single-point fault is that all systems are all downtime when a component or an application fails in the system. In this system, the disk array cabinet will result in single point fault. When the disk array cabinet is logical or physical failure, all stored data will be all lost, so when you choose this, you need to choose a quality and after-sales. Good service products.
Hardware configuration example:
• Two network servers
• Server operating system hard disk two pieces
• One disk array cabinet produced by third parties
• Two-purpose SCSI wires for disk cabinets
• Disk Array Cabinet Data Storage Hard Disk is determined by user requirements
• Two blocks of network service network card
In addition, some manufacturers have better technology solutions, such as HP dual-machine dual-control fault tolerant systems, HP Fiber Channel Dual Control Cluster System.
3, cluster software configuration
Cluster software based on NT platform
Microsoft's MSCs, there are also many third-party professional software companies developed cluster software, such as Haowa's DataWare, Vinca's Standby Server, NSI's Double-Take.
MS WolfPack is the other name of MS Cluster Server, which is a dual-machine software developed by Microsoft to develop developments in Cluster technology. It is integrated on NT Server, supports two-machine system consisting of two machines, providing a highly available and manageable application environment.
main feature:
• Automatically detect and repair errors or applications
• Switching to the application in the server can be implemented
• Various clients can be connected via TCP / IP, such as MS-DOS, Windows 3.x / 9x / NT, Apple Macintosh, UNIX, etc.
• Production hosts can automatically recover data and take over tasks without manual interference
• Easy management:
• Automatically audit the work status of servers and applications
• Establish high availability applications, file sharing, print request, etc.
• Flexible settings of the application and data recovery strategy
• Simple operations can be offline, re-line, and migration between servers.
Currently, Windows 2000 Advanced Server and Windows 2000 DataCenter Server are integrated with more advanced cluster technology.
There are also many cluster software on other network operating system platforms, such as:
Cluster software based on Novell platform has Novell HA Server, Novell SFT III
Cluster software based on SCO UNIX platform has Sentinel cluster software
Cluster software based on Linux platform has TurboCluster
4. Development trend of cluster technology
Cluster technology will gradually increase in the development of server hardware systems and network operating systems. Future clusters can rely on cluster file systems to implement global access to all files, devices, and network resources in the system, and generate a complete system image. In this way, the cluster file system allows any user (remote or local) to access this software on any user (remote or local) on which server is in the cluster. Any application can access any files in this cluster. Even if the application is transferred from a node to another, the application can access the file on the system. Today, using the server's cluster technology, the system destruction is very small through careful plans and network maintenance. Therefore, the virtual host uses cluster technology will be quite stable.
