The overview and definition of clusters can be seen at a glance

Hello everyone , Today I share an overview and definition of cluster

Our first question is ：

What is cluster ：

Put many servers together , Make them a whole , By a unified machine （ The server ） Conduct management , Realization Processing a large amount of data information

Here are the results of Baidu

A cluster is a group of independent 、 A computer connected by a high-speed network , They form a group , And managed in a single system model . When a customer interacts with a cluster , Clustering is like a stand-alone server . Cluster configuration is used to improve availability and scalability .

My vernacular is
The scheduler directs this pile of independent computers to do one thing , Such a system , It's clustering

About the characteristics of clusters

Sure

At low cost , Realize the performance of the server , reliability , Flexibility （ It is also the purpose of cluster ）

The core of clustering is Scheduler

Purpose of cluster

1. Lies in the improvement of performance （ That is, computing power and the ability to handle high concurrent data traffic ）
2. lie in Cost reduction （ because , Compared with the high price of supercomputers , The cluster built by traditional servers greatly reduces the cost ）
3. It can improve scalability （ Namely , If you need to change , Add or delete the server node directly ）

The feeling of cluster for users is ：（ When the user receives the visual interface ）

Whether in the cluster processing requests , How many servers are used , user It will be considered that a server is actually working

There is a classification of clusters

1. High performance computing cluster （HPC）, Through cluster development , Allow applications to appear in parallel , Mainly to solve a large number of complex scientific problems

2. Load balancing cluster （LB）, That is to say , In such a cluster , What is dealt with Data and information flow of the client , Try to give it to the real working servers in the cluster in an equal distribution way

For example ,100G The flow of is distributed to 10 Servers , Each server handles 10g Of traffic
3. High availability cluster （HA）, It is in such a cluster , An unexpected situation occurred on a single server , It can realize rapid migration and backup , The purpose is to avoid single point of failure

LVS Introduce
lvs , yes linux Short for virtual server , It was founded by Dr. zhangwensong during his study

lvs It can achieve high availability and load balancing

yes Linux Functions and services embedded in the system

About lvs The architecture of the figure ：

Its purpose is to make use of linux and lvs High availability , High performance , Low cost cluster environment

lvs Component part
front end ： Load balancing layer
It is a layer composed of one or more schedulers

The one I drew , It's the scheduler

middle ： It's a bunch of real servers

Back end ：
The layer responsible for data sharing （ In the picture below ）

stay lvs In the cluster , There are also some name terms

1. Directoy : This is the scheduler （ The core of the cluster ）
2. real server ： Real server （ Really deal with data , Servers that really work ）
3. VIP : For users to access IP Address
4. RIP : real ip Address （ The server node is used to communicate with other nodes in the cluster ip Address ）
5. DIP : The scheduler connects to the node server ip Address

One more ： CIP ( Client's ip Address , This is not within the jurisdiction of the cluster )

Okay , Come here , We have an overview and definition of cluster, that's all , Thank you.