Distributed cluster architecture scenario optimization solution: session sharing problem

Distributed cluster architecture scenario optimization solution ：Session Sharing issues

Session Sharing issues

Session Sharing and Session Keep or call Session Uniformity

1、 Session Problem cause analysis

The reason for this problem , Basically, it's because Http Protocols are stateless protocols . The data generated by the client and server in a session will not be preserved , So the second request server can't realize that you've been here , Http Why should it be designed as a stateless protocol ？ In the early days, it was static pages, and it didn't matter whether there was a state or not , Later, there are more dynamic contents , You need to be in a state , There are two ways to keep Http Technology of state , That's it Cookie and Session. And the above problem of keeping logging in , The analysis is shown in the figure below ：

scene ：Nginx Default polling policy

2、 solve Session A consistent solution

1、Nginx Of IP_Hash Strategy （ have access to ）

Same client IP All requests will be routed to the same target server , Also called session stickiness

advantage ：

• Simple configuration , Do not invade the application , No additional code changes are required

shortcoming ：

• Server restart Session The loss of
• There is a risk of high single point load
• Single point problem

2、Session Copy （ Not recommended ）

Multiple tomcat By modifying the configuration file , achieve Session Copy between

advantage ：

• Do not invade the application
• It is convenient for the horizontal expansion of the server
• It can adapt to various load balancing strategies
• Server restart or downtime will not cause Session The loss of

shortcoming ：

• Low performance
• Memory consumption
• Can't store too much data , Otherwise, the more data, the more performance will be affected
• Delay

3、Session share ,Session Centralized storage （ recommend ）

Session The essence of is caching , that Session Why not hand over the data to professional caching middleware ？ such as Redis

advantage :

• It can adapt to various load balancing strategies
• Server restart or downtime will not cause Session The loss of
• Strong expansion ability
• Suitable for large number of clusters

shortcoming ：

• Intrusion into applications , And are introduced Redis Interactive code

Spring Session Make based on Redis Of Session Sharing applications is very simple

1） introduce Jar

<dependency>
	<groupId>org.springframework.boot</groupId>
	<artifactId>spring-boot-starter-data-redis</artifactId>
</dependency>
<dependency>
	<groupId>org.springframework.session</groupId>
	<artifactId>spring-session-data-redis</artifactId>
</dependency>

2） To configure redis

spring.redis.database=0
spring.redis.host=127.0.0.1
spring.redis.port=6379

3） adding annotations

3、 Source code

Observe its parent class , In the parent class Filter

This Filter Namely SpringSession At the core

In the filter HttpServletRequest packing

It's essentially a HtppRequest, Have the same approach , look for getSession

go back to SessionRepositoryFilter Of doFilterInternal Method

Schematic diagram