当前位置:网站首页>Proteus is a new platform for building streaming microservices using rsocket

Proteus is a new platform for building streaming microservices using rsocket

2021-05-04 14:01:42 Jiedao jdon

Netifi Proteus Is similar to Spring Cloud Of , But based on RSocket Build the next generation of reactive micro service platform , It allows developers to transparently provide API management , route , Service discovery , Predict load balancing and ultra low latency RPC To focus on their products .

Spring Cloud Communication between components is usually based on Pull Of Http Way of agreement , and Proteus It's based on push RScoket + Protobuf Dynamic mode .

Proteus = SpringCloud + Vert.x

Proteus Adopt a novel method to solve the problem of network efficient communication 、 Circuit breakers and load balancing , Instead of layering over traditional communication protocols .Proteus Provides a solution based on three layers :

  1. Proteus It's reactive , Not only in the application layer , And at the network level . Unlike other solutions built on traditional network models ,Proteus Built on an application network stack designed for cloud native applications . This gives you a powerful tool for building real-time streaming applications .
  2. Proteus Very clever . Unlike other solutions that offer more settings and require frequent manual adjustments ,Proteus Use information about your application and application traffic to prevent thunderous crowding , Retrying storms and other frequently asked questions .
  3. Proteus Provide strong contracts between applications   The contract provides you with implicit version information , Information about exchanging data , Interaction model between applications and state aware service discovery .

Message reorganization and back pressure aware flow

Whether the application is sending short HTTP Request or megabyte streaming video ,Proteus The message routing provided is very efficient .RSocket With built-in message blocking / Refactoring logic , Manageable payload size enables streaming large data sets .

Proteus Support various back pressure strategies , To make sure that data consumers can keep up with data producers , No one needs to implement their own back pressure or retrial logic . As a service , You can set up a mechanism to stream all the data clients may need , And the demand and consumption rate of the client can adjust the actual speed of data transmission through the line - Instead of any complex server-side logic .

Multilingual support

client , Servers and peer-to-peer applications can be written in multiple languages , And then you can use RSocket Open source protocol for full duplex , Reactive communication ,Protobuf  Interface guarantees mutual compatibility definition language .

The ecological system

Proteus The main components are Proteus Broker, It's a message bus like agent , Its complete ecosystem includes :

RSocket RPC

Support Protobuf And primitive RSocket The reaction of RPC layer . You can do it with or without Broker Use it in case of .

Proteus Broker

Cloud native message broker Broker, Provides virtual connectivity directly between applications , Wherever it is .Proteus Agents facilitate communication between client and server applications - Including mobile devices ,Web Browsers and microservices in the cloud - No matter where they are , It's like they have point-to-point connections .

Proteus Agents are built on Reactive Streams and RSocket above . It's a non blocking application , It creates the optimal number of threads based on the underlying transport ( for example , Use TCP and WebSockets when , It will create a thread for each core ).

Proteus No need to use long polling or periodic queries , To understand which client and service applications are active . contrary , It depends on RSocket Connection oriented features of , And as long as the application and Proteus The proxy remains connected , Just keep information about each application .Proteus The agent shares this connection information with any other agent in its cluster , So that state changes can be propagated to other agents in time .

Borker colony

Multiple Broker We can work together , Improve reliability and scalability .

Proteus Console

Proteus Portal for , Provides management and monitoring .

Proteus CLI

Manage and ask from the command line Proteus The state of the agent cluster .

Proteus client

You can achieve through Proteus The client and microservice that the system communicates with .

Proteus HTTP gateway

Using agents and Proteus RPC Create... For legacy applications Proteus Ecosystem HTTP / 1.1 Restful Bridged HTTP gateway .

Proteus indicators

Allows you to Proteus The application collects metrics , And connect them to the existing indicator collection system .

Proteus Open Tracing

utilize Proteus Ecosystems and collect from anywhere Open Tracing Information .

https://github.com/netifi/proteus-spring-kafka-example  Shows how to use Proteus and reactor-kafka Realize message integration .

https://github.com/netifi/RxNetty yes Netty Responsive expansion adapter for .

版权声明
本文为[Jiedao jdon]所创,转载请带上原文链接,感谢
https://chowdera.com/2021/05/20210504135746267m.html

随机推荐