Linux server cluster system implementation solution

I, Cluster Basic Concepts

There is a common way to greatly improve the security of the server, which is the cluster.

The cluster technology can be defined as follows: a group of independent servers act as a single system in the network and are managed in a single system mode. This single system provides high reliability services for the customer workstation.

In most modes, all the computers in the cluster share a common name. services running on any system in the cluster can be used by all network customers. The cluster must be able to coordinate and manage the errors and failures of the separated components, and add components to the cluster transparently.

A cluster contains multiple (at least two) servers with shared data storage space. When any server runs an application, application data is stored in the shared data space. The Operating System and Application Files of each server are stored in their local storage space.

Each node server in the cluster communicates with each other through an internal LAN. When a node server fails, the applications running on this server will be automatically taken over on another node server. When an application service fails, the application service will be restarted or taken over by another server. When any of the above faults occurs, the customer will be able to quickly connect to the new application service.

　　Ii. Cluster hardware configuration

Backup storage

In a cluster, the dual-host backup storage system is the simplest and cheapest solution for hardware configuration. Generally, the hardware configuration of the backup storage service requires two servers, each server has an independent operating system hard disk and data storage hard disk. Each server has a network card connected to the client, and a pair of mirror cards or network cards that complete the image function.

The backup storage has many advantages, such as simple configuration, easy to use, and low price. However, the backup storage must use network-based backup storage to synchronize data, therefore, the CPU and memory resources of the network server need to be occupied. The performance of the backup storage is lower than that of a single server.

Some image server cluster systems use the memory image technology. The advantage of this technology is that all applications and network operating systems synchronize images on the two servers. When the host fails, the backup machine can take over all applications with almost no feeling. Because the memory of the two servers is completely the same, but when the system application program has a defect that causes the system to go down, the two servers will go down simultaneously. This is also the memory image card or network card for data synchronization. In the process of reading and writing large data volumes, the two servers may not synchronize data in some statuses, therefore, the backup storage is suitable for users with low budget and low requirements on Cluster Systems.

Hardware configuration example:

Network server, two operating system disks

Server Data Storage hard disk depends on user needs

Server image card (some software can use standard Nic)

Two network service NICs: Dual-host and disk array

Compared with the dual-host system of the backup storage, the interconnection structure between the dual-host and the disk array is larger than that of the third-party disk array cabinet. Currently, many companies such as haowei and Jingye produce disk array cabinets, the disk array control card is installed in the disk array cabinet. The array cabinet can directly configure the hard disk in the Cabinet as a logical disk array. The disk array cabinet is connected to a common scsicard on the server through the SCSI cable. The system administrator must directly configure the disk array on the cabinet.

The interconnection structure between the dual-host and the disk array does not use the memory image technology, so a certain switching time is required (usually 60? D? D180 seconds), it can avoid all system downtime due to application defects, and because all data is stored in the middle disk array, when the working machine fails, the backup machine takes over the working machine and reads data from the disk array. Therefore, data is not synchronized because this solution does not require network image synchronization, therefore, the performance of this cluster solution server is much higher than that of the backup storage.

The disadvantage of the dual-host and disk array interconnection structure is that there is a single point of failure in the system. The so-called single point of failure refers to when a part or an application in the system fails, all systems are down. In this system, a disk array causes a single point of failure. When a logical or physical fault occurs in the disk array, all the stored data is lost. Therefore, when selecting this solution, select a product with good quality and after-sales services.

Hardware configuration example:

Two network servers

Two server operating system hard disks

One third-party disk array Cabinet

Two dedicated SCSI wires for the cabinet

Disk Array cabinet data storage hard disk determined by user requirements

Two network service NICs

In addition, some vendors also have better technical solutions, such as HP.

Iv. HP dual-host dual-controller fault toleranceSystem

HP NETServer provides a high quality and highly reliable hardware foundation for dual-host dual-controller fault tolerance systems ......

The HP dual-host dual-controller fault tolerance system combines the security and reliability of HP Server products with the advantages of cluster technology, and cooperates with each other to provide the advantages of both.

Hardware configuration example:

Two network servers of the hp l System

Two server operating system hard disks

One HP hard disk storage cabinet (SS/6, RS/8, RS/12)

Dedicated SCSI cluster adapter cable for Cabinet

Cabinet data storage hard disk determined based on user requirements

Two dedicated array cards for HP Clusters

Network Service Network Card Module 5, HP Fiber Channel dual-host dual-controller Cluster System

Fiber Channel is a connection Standard and can be used as an alternative solution for SCSI, optical fiber technology has high bandwidth, anti-electromagnetic interference, long distance transmission, high quality, strong scalability and other characteristics, currently in the FC-AL arbitration loop can access 126 devices.

Fiber optic devices provide a variety of enhanced connection technologies, greatly facilitating your use. The server system can be remotely connected through optical cables, up to 10 kilometers away. It allows image configuration to improve the fault tolerance capability of the system. The scale of the server system will be more flexible and variable. Each SCSI channel can connect up to 15 devices, and the optical fiber arbitration loop can connect up to 126 devices.

Composition of the optical fiber cluster system:

The hardware of the HP Optical Fiber cluster system includes two HP servers (including lc2000, lh3000, lh4, lh6000, lt6000, lxr8000, and lxr8500) and Optical Fiber adapter, you can use the RS/12fc Optical Fiber disk array cabinet. You need to add one or two pairs of NICs for heartbeat detection and connection to the client. During the configuration process, you also need to select an optical fiber cable from the optical fiber card to the optical fiber storage device. Hardware configuration:

Two hpl system network servers

Two server operating system hard disks

One HP optical fiber array storage cabinet (RS/12fc)

Two optical fiber cables for optical fiber Cabinet

Optical fiber cabinet data storage hard disk determined by user requirements

HP Optical Fiber adapter

Two network service NICs
VI,ClusterSoftware Configuration

NT-based cluster Software

Microsoft's MSC also has cluster software developed by many third-party professional software companies, such as haowei's dataware, Vin CA's standby server, and NSI's double-take

Features of MS wolfpack

MS wolfpack is the nickname of MS cluster server, which is the dual-machine software developed by Microsoft for cluster technology. It is integrated on the NT Server and supports a dual-machine system composed of two machines to provide a highly available and manageable application environment.

Main features:

Automatically detect and fix server or application errors

Allows you to switch between applications on the server.

You can use TCP/IP to connect to various clients, such as MS-DOS and Windows 3.x/ 9x/NT. Apple Macintosh and Unix production hosts can automatically restore data and take over tasks without manual interference.

Easy to manage:

Automatically reviews the working status of servers and Applications

High-availability applications, file sharing, and print requests can be created.

Flexible setting of recovery policies for applications and data

With simple operations, you can perform offline, rewiring, and migration between servers.

Currently, Windows 2000 Advanced Server and Windows 2000 datacenter server are integrated with more advanced cluster technologies.

There are also many cluster software on other network operating system platforms, such:

The cluster software based on the Novell platform includes Novell ha server and Novell sft iii.

The cluster Software Based on sco unix platform includes sentinel cluster software.

The Linux-based cluster software includes turbocluster.

　　VII. Development Trend of cluster technology

With the development of server hardware systems and network operating systems, cluster technology will gradually improve in terms of availability, high reliability, and system redundancy. In the future, clusters can use the Cluster File System to globally access all files, devices, and network resources in the system and generate a complete system image. In this way, regardless of the server in the cluster, the cluster file system allows any user (remote or local) to access the software. Any application can access any files in the cluster. Even when an application is transferred from one node to another, the application can access files on the system without any changes.

Today, using the cluster technology of servers, through careful planning and network maintenance, the probability of system damage is very small. Therefore, cluster technology must be used to ensure the stability of enterprise servers.