Distributed --zookeeper

 

The number of visits that the system can receive in a limited time is an important criterion for evaluating the success of a system for the general public. In the traditional scheme, the way to increase the number of visits is to increase the number of servers. For example, if a server can support 100 people accessing at the same time, and there are 10,000 people accessing the system at the same time, then 100 servers are provided. However, in the traditional method, the administrator project and the user project are placed on the same server, which is a great waste of resources. Obviously, the number of administrators in a system is far less than the number of users. It is unnecessary to place the administrator system on most servers. At this time, it is necessary to adopt a distributed system. As shown in the figure below, the administrator system and the user system are placed separately, and the two share data through the zookeeper registration center.

Features of Zookeeper

1 Eventual consistency : Presenting the same view to clients is the most important function of zookeeper.
2 Reliability : If a message is accepted by one server, then it will be accepted by all servers.
3 Real-time : Zookeeper cannot guarantee that two clients can get the newly updated data at the same time.
4 Wait -free: slow or failed clients do not interfere with fast client requests.
5 Atomicity : Updates can only succeed or fail, with no intermediate states.
6 Sequentiality : All servers, the same message is published in the same order.

 

Guess you like

Origin http://43.154.161.224:23101/article/api/json?id=325729130&siteId=291194637