What Is Application Clustering?

A server cluster is a collection of many servers to perform the same service together. It appears to the client that there is only one server. The cluster can use multiple computers for parallel computing to obtain high computing speed, and it can also use multiple computers for backup, so that any one machine can still run normally if the entire system is broken.

Server cluster

Once the cluster service is installed and running on the server, the server can join the cluster. Clustering operations can reduce the number of single points of failure and achieve high availability of clustered resources. The following sections briefly describe node behavior in cluster creation and cluster operations.

Note: About installation

in

If the server is running the cluster service and cannot find other nodes in the cluster, it can form a cluster by itself. To form a cluster, nodes must be able to obtain

From the perspective of the other nodes in the cluster and the cluster service management interface, when a cluster is formed, each node in the cluster may be in one of three different states. The event handler records these states, and the event log manager replicates these states to other nodes in the cluster. The cluster service status includes:

Offline. The node at this time is not a fully valid cluster member. The node and its

First, the cluster system can resolve all server hardware failures. When any failure occurs on a server, such as: hard disk, memory, CPU, motherboard, I / O board and power failure, applications running on this server will switch Go to another server.

Second, the cluster system can solve software system problems. We know that in computer systems, users use applications and data, and the application system runs on the operating system, and the operating system runs on the server. In this way, as long as any one of the application system, operating system, and server fails, the system actually stops providing services to the client, such as our common software

We know that the applications in the cluster only run on one server. If this application fails, some other server will restart the application and take over the data area on the shared disk cabinet, so that the application can run normally again. We know that the entire application takeover process generally requires three steps: detecting and confirming the failure, the backup server restarting the application, and taking over the shared data area. Therefore, it takes a certain time during the switching process. In principle, the switching time will be different according to the size of the application. The larger the application, the longer the switching time.

When a node is shut down or the cluster service is stopped, the node may leave the cluster. But when a node does not perform cluster operations (such as not submitting updates to the cluster configuration database), the node may also be forced to leave (evicted) the cluster.

If a node leaves the cluster according to a predetermined schedule, it sends a ClusterExit message to all other node members informing them that it will leave the cluster. The node immediately shuts down resources and all cluster connections without waiting for any response. Since the rest

There are two commonly used server cluster methods. One is to connect the backup server to the master server. When the master server fails, the backup server is put into operation and takes over all tasks on the master server. Another method is to connect multiple servers, which share the same application and database computing tasks together to improve the response time of key large applications. At the same time, each server also assumes some fault tolerance tasks.Once a server fails, the system can

In a cluster system, all computers have a common name, and services running on any system in the cluster can be used by all network customers. The cluster must be able to coordinate and manage errors and failures of separate components, and transparently add components to the cluster. The user's public data is placed in a shared disk cabinet, and the application is installed on all servers, that is, the application running on the cluster needs to be installed on all servers again. When the cluster system is running normally, the application runs on only one server, and only this server can manipulate the data area of the application on the shared disk cabinet.

1. High availability cluster

   The full English name of a high availability cluster is High Availability, referred to as HA cluster. High availability means maximum use. As can be seen from the name of the cluster, the function implemented by this cluster is to ensure that the user's application provides services persistently and uninterruptedly.

2. Load balancing cluster

   A load balancing cluster is also composed of two or more servers. It is divided into two parts: front-end load scheduling and back-end service. The load scheduling part is responsible for distributing client requests to back-end service nodes according to different policies, and the back-end nodes are the parts that truly provide application services. Unlike HA Cluster, in a load balancing cluster, all back-end nodes are active, they all provide services to the outside world, and share the workload of the system.

3. Scientific Computing Cluster

   High-performance computing cluster, referred to as HPC cluster. This type of cluster is dedicated to providing powerful computing capabilities that cannot be provided by a single computer, including numerical calculations and data processing, and tends to pursue comprehensive performance. HPC is similar to, but different from, supercomputing. Computational speed is the first goal pursued by supercomputing. The fastest speed, the largest storage, the largest volume, and the most expensive price represent the characteristics of supercomputing. With the increase of people's demand for computing speed, supercomputing is also applied to various fields. The pursuit of a single computing speed indicator for supercomputing has turned into a comprehensive indicator for high performance, that is, high performance computing. ^[1]