High-availability distributed micro-service architecture

Abstract

The invention discloses a high-availability distributed micro-service architecture, which comprises middleware, a database and a storage system, and comprises a Master node, a StorageNode data node and a Client node, and the Master node completes a task of reading and writing from a distributed file system with a central control node; the StorageNode data node is used for storing files on a stand-alone system, and for storage nodes with master and slave differences, backup data consistency is kept; the Client node is an access layer, performs write operation, receives user data flow and writes data into the storage node, and the middleware can be dynamically expanded. According to the invention, at least three nodes are served to realize the load balancing of the service, and the effects of automatically deploying each node and automatically expanding the capacity are realized.

Description

technical field [0001] The invention relates to the technical field of data storage, in particular to a highly available distributed micro-service architecture. Background technique [0002] The high-availability distributed micro-service architecture of the prior art includes the connection board nacos cluster, rocketmq cluster, and emqx cluster. The current services in the prior art are all deployed in a cluster mode, support high availability, load in a load balancing mode, and realize automatic capacity expansion and automatic restart. Use monitoring to monitor the current load situation of the cluster, and realize automatic restart and automatic expansion. The program deployment method uses the open source jenkins to realize automatic half-gray release, automatic release to realize the health check of services, and automatic detection of service ports. [0003] The architecture of the existing technology has a wide range of failures and poor scalability. The entire sy...

AI Technical Summary

Problems solved by technology

[0003] The architecture of the existing technology has a wide range of failures and poor scalability. The entire system will be unavailable due to a machine failure. Therefore, it is necessary to use a distributed architecture to redundant the system to eliminate single points of failure, thereby improving the system. availability of

[0004] Moreover, the existing technology has the following technical problems, 1. The architecture design is complex (especially the distributed transaction); 2. It is faster to deploy a single service, but if multiple services are required for one deployment, the deployment will become complicated; 3. , The throughput of the system becomes larger, but the response time will be longer

4. Increased complexity of testing and troubleshooting

Images