A brief description of how Fourinone distributed and coordinated design

Distributed collaboration is indispensable in distributed application, and it is the role of coordinator, that is, to separate the responsibility of multi-machine synergy from distributed application, so as to reduce the coupling of system and enhance the expansibility. The zookeeper of Apache, Google's chubby are the implementation of distributed collaboration. Fourinone is actually an upgraded version of Zookeeper, which implements all of zookeeper's functionality with minimal code and strives to use more concise, more powerful applications.

First, the realization principle

Fourinone to the implementation of distributed collaboration, through the establishment of a domain,node two-tier structure of the node information to complete, domain can be classified or package, node can be specific attributes, domain and node are themselves according to the requirements of the design name, For example, you can name domain "A.B.C ..." Represents a tree type.

A domain can have many node, each node only specify a domain, you can return it through domain all the node below it.

Domain does not need to be created separately, usually when you create node, if domain is not present automatically.

If there is no node in domain, the domain is automatically deleted.

If you delete domain, node under this domain will also be deleted.

A value can be stored under each node, and can be any object.

All the node information is stored in the Parkserver, Parkserver provides the function of the cooperating person. As shown in the following illustration:

As can be seen from the above diagram, other distributed processes can add, modify, delete, specify heartbeat, assign permissions and other functions through Parkserver user interface parklocal, and combine parkserver to provide synchronous backup, leader election, expiration time setting, etc. Together to achieve a number of distributed synergy functions, such as:

1, distributed configuration, the application of multiple machines a common configuration information, and hang out to be able to lead the election, detailed see guide and demo

2, distributed lock, multiple machines to compete for a lock, when a machine release lock or hang, other machines can compete to lock continue, detailed see Guide and demo

3, cluster management, clusters within the machine can be perceived and leaders of the election, see the Guide and demo