当前位置:网站首页>Microservice registration and discovery
Microservice registration and discovery
2022-07-06 02:41:00 【Rainy night】
Multi service communication under micro service architecture , One problem that needs to be solved is how to realize service discovery . This article will start with why service discovery is needed 、 What is service discovery 、 How to realize service discovery is briefly introduced in four aspects .
Why service discovery is needed
When calling between services , One service needs to know the network location of another service (IP Address and port ). Applications running on physical hardware , The network location of the service instance is usually static . But in the cloud age , It's usually not that simple . In cloud based service instances , Its network location is more dynamic . An example is shown below :
Pictured above is an example , When creating service instances based on cloud , Service instances have dynamically allocated IP Address . Because service instances can be created or destroyed at any time , Therefore, a service instance cannot occupy a fixed IP, Instead, a random one is used every time a service is created IP, Recycle this when the service is destroyed IP.
What is service discovery
The service caller cannot configure IP Address to access the service provider interface . contrary , Applications or infrastructures need to provide dynamic service discovery mechanisms . Service discovery is conceptually very simple : The key component of service discovery is the service registry , It is a database containing the network location information of service instances .
When the service instance starts or stops , The service discovery mechanism updates the service registry . When the client calls the server , The service discovery mechanism will query the service registry , To get a list of available service instances , And route the request to one of the service instances . Careful students can find out , This principle is similar to DNS similar . Actually , The principle of service discovery is to learn DNS Principle realization .
How to implement service discovery
There are two main ways to implement service discovery :
(1) Service callers and service providers interact directly with the service registry , That is, the application layer service discovery pattern .
(2) Handle service discovery by deploying infrastructure , That is, the platform layer service discovery pattern .
Application layer service discovery pattern
One way to realize service discovery is that service callers and service providers interact directly with the service registry . The service provider registers its network location using the service registry . The service caller first calls the service by querying the service registry to obtain the list of service instances , Then send a request to one of the instances . Examples are as follows :
It can be seen from the figure above , When initializing the service instance , Will call registration API Register its network location with the service registry . in addition , The service registry may also require service instances to call periodically heartbeat API To prevent its registration from expiring . When the service caller needs to call the service interface , It queries the service registry for a list of service instances . Then the service caller uses Load balancing algorithm To select a service instance . Then it sends a request to the selected service instance .
In the early stage of the development of microservice architecture , The application layer service discovery model has been greatly promoted . The main application layer service discovery components are Netflix Of Eureka. This is also Spring Cloud The default service discovery component before the framework (Eureka Maintenance is now stopped ).
One advantage of application layer service discovery is that it can handle the problem of multi platform deployment ( The service discovery mechanism is independent of the underlying deployment platform ).
One disadvantage of application layer service discovery is that each service discovery library is language or framework related . Such as Eureka Can only help based on Java Developer . If needed GoLang Or other languages , You must use other service discovery components .
Platform layer service discovery pattern
Another way to implement service discovery is to deploy infrastructure to handle service discovery . Many modern deployment platforms ( Such as Docker and Kubernetes) Both have built-in service registry and service discovery mechanism . The deployment platform provides services for each service DNS name 、 fictitious IP(VIP) And resolve to VIP Address of the DNS name . Service call direction DNS Name and VIP Request , The deployment platform automatically routes requests to one of the available service instances . therefore , Service registration 、 Service discovery and request routing are handled entirely by the deployment platform . Examples are as follows :
The deployment platform includes a service registry , Used to track deployed services IP Address . When the service starts , The Registrar of the deployment platform writes the address of the service instance into the registry . Each service has a network location 、 One DNS Name and virtual IP Address . The service call requests the network location of the service provider from the router of the deployment platform . Then deploy the router of the platform to query the registry and load balance requests between available service instances .
The main advantage of the service discovery mechanism provided by the platform is that all aspects of service discovery are completely handled by the deployment platform . Neither the service provider nor the service caller contains any service discovery code . therefore , No matter which language or framework you use , The service discovery mechanism can be used by all service providers and service callers .
One disadvantage of the service discovery mechanism provided by the platform is that it is limited to services that support deployment using the platform . Based on Kubernets The findings of apply only to Kubernets Services running on .
From the perspective of the development of microservice architecture , The platform layer service discovery model has been successfully mainstream in the industry , Cloud service development I participated in , It is based on the platform layer service discovery pattern .
Reference resources
Microservice design Sam Newman Writing , Cui Liqiang etc. translate
Microservice architecture design pattern Chris Richardson Writing , Chen Bin etc. translate
https://itdks.su.bcebos.com/307ea09e7ef34bbfa58b845bf0f3d74b.pdf Microservice architecture development and platform evolution
边栏推荐
- Gifcam v7.0 minimalist GIF animation recording tool Chinese single file version
- HDU_p1237_简单计算器_stack
- [Yunju entrepreneurial foundation notes] Chapter II entrepreneur test 9
- 2.12 simulation
- Introduction to robotframework (I) brief introduction and use
- Force buckle 146 LRU cache
- Thinking on Architecture Design (under continuous updating)
- Bigder:34/100 面试感觉挺好的,没有收到录取
- How to read excel, PDF and JSON files in R language?
- Pure QT version of Chinese chess: realize two-man, man-machine and network games
猜你喜欢
[Yunju entrepreneurial foundation notes] Chapter II entrepreneur test 16
2345文件粉碎,文件强力删除工具无捆绑纯净提取版
全国大学生信息安全赛创新实践赛初赛---misc(永恒的夜)
Shell script updates stored procedure to database
从顶会论文看2022年推荐系统序列建模的趋势
Easy to use js script
[Yunju entrepreneurial foundation notes] Chapter II entrepreneur test 18
[Yunju entrepreneurial foundation notes] Chapter II entrepreneur test 13
解决:AttributeError: ‘str‘ object has no attribute ‘decode‘
Introduction to robotframework (II) app startup of appui automation
随机推荐
[Wu Enda machine learning] week5 programming assignment EX4 - neural network learning
2.11 simulation summary
Large scale DDoS attacks take Myanmar offline
Gifcam v7.0 minimalist GIF animation recording tool Chinese single file version
MySQL winter vacation self-study 2022 11 (8)
深度解析链动2+1模式,颠覆传统卖货思维?
[Yunju entrepreneurial foundation notes] Chapter II entrepreneur test 20
Microsoft speech synthesis assistant v1.3 text to speech tool, real speech AI generator
C language - Blue Bridge Cup - promised score
零基础自学STM32-野火——GPIO复习篇——使用绝对地址操作GPIO
会员积分营销系统操作的时候怎样提升消费者的积极性?
What should we pay attention to when using the built-in tool to check the health status in gbase 8C database?
Number conclusion LC skimming review - 1
"Hands on learning in depth" Chapter 2 - preparatory knowledge_ 2.5 automatic differentiation_ Learning thinking and exercise answers
2020.02.11
A doctor's 22 years in Huawei
模板_求排列逆序对_基于归并排序
微服务注册与发现
MySQL winter vacation self-study 2022 11 (7)
[Yunju entrepreneurial foundation notes] Chapter II entrepreneur test 23