Remember once springcloud concurrent optimization

Using components springcloud kit + mysql

A common request of rest, to be a basic user information query, sql been optimized const level, but the pressure-measurement by jemeter, found that the actual amount of concurrency only 20-30, after inspection found that the configuration is not optimized problem solution is as follows:

 

1. Increase springboot mysql connection pool configuration, HikariDataSource to use by default. Configuration is as follows

spring:
  datasource:
    type: com.zaxxer.hikari.HikariDataSource
    driver-class-name: com.mysql.cj.jdbc.Driver
    username: cjs-base
    password: xxx
    url: jdbc:mysql://xxx:3306/xxx?characterEncoding=utf8&zeroDateTimeBehavior=convertToNull&useSSL=false&allowPublicKeyRetrieval=true
    hikari:
      minimum-idle: 100
      maximum-pool-size: 500
      idle-timeout: 30000
      max-lifetime: 180000
      # connection-timeout: 5000
      Connection -test-Query: SELECT  . 1 
      Connection -init-SQL: SELECT  2 
      # how many milliseconds detecting a leaking connection. 
      Leak -detection-threshold: 1000 
      Test -ON-BORROW: to true 
      Test -ON-Connect: to true 
      Test -On- return : to true 
      Test - the while -idle: to true 
      # specify idle connection check, cleaning waste connection, connection idle operating time interval between the pool size adjustment 
      time -between-eviction-the runs of millis-: 10000 
      # whether connection packaging, to prevent the connection closed after being used. 
      use -disposable-Connection-Facade:to true 
      Validation -interval: 10000 
      alidation -query: SELECT  . 3 
      Validation -query-timeout: 5000 
      # connection check timeout setting, specifies the pool when a Connection Hikari 
      alidation -timeout: 5000    

In which the connection pool size, configuration items according to their own situation.

 

2. Modify fegin configuration, the general default is like. Mainly to prevent resource consumption can not be released, we need to set a relatively small timeout.

feign:
  hystrix:
    enabled: true
  okhttp:
    enabled: true
  httpclient:
    enabled: false
  client:
    config:
      feignName:
        connectTimeout: 1000
        readTimeout: 1000
  compression:
    request:
      enabled: true
    response:
      enabled: true

　3. Modify hystrix configuration, the connector size is set to be greater than the amount required concurrent

hystrix:
  threadpool:
    default:
      coreSize: 500
      maxQueueSize: 1000
  command:
    default:
      circuitBreaker: 
        requestVolumeThreshold: 50
      execution:
        isolation:
          semaphore:
            maxConcurrentRequests: 1000
          strategy: SEMAPHORE
          thread:
            timeoutInMilliseconds: 1000
  shareSecurityContext: true

　　4. Modify ribbon configuration

# Timeout process request 
Ribbon: 
  ReadTimeout: 1000 
  the ConnectTimeout: 1000

The increased number of instances of micro and services, expansion in the number of each micro-demand services K8S, each micro my side expansion three service instances, with the above configuration optimization, concurrency can be controlled at about 250 It has to meet current business needs. Such as the need for further optimization, can be used to increase the service instance, increased to achieve the implementation of NOSQL