zoukankan      html  css  js  c++  java
  • HAProxy advanced Redis health check---ref

    http://blog.exceliance.fr/2014/01/02/haproxy-advanced-redis-health-check/

    HAProxy advanced Redis health check

    Introduction

    Redis is an opensource nosql database working on a key/value model.
    One interesting feature in Redis is that it is able to write data to disk as well as a master can synchronize many slaves.

    HAProxy can load-balance Redis servers with no issues at all.
    There is even a built-in health check for redis in HAProxy.
    Unfortunately, there was no easy way for HAProxy to detect the status of a redis server: master or slave node. Hence people usually hacks this part of the architecture.

    As written in the title of this post, we’ll learn today how to make a simple Redisinfrastructure thanks to newest HAProxy advanced send/expect health checks.
    This feature is available in HAProxy 1.5-dev20 and above.

    Purpose is to make the redis infrastructure as simple as possible and ease fail over for the web servers. HAProxy will have to detect which node is MASTER and route all the connection to it.

    Redis high availability diagram with HAProxy

    Below, an ascii art diagram of HAProxy load-balancing Redis servers:

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    +----+ +----+ +----+ +----+
    | W1 | | W2 | | W3 | | W4 |   Web application servers
    +----+ +----+ +----+ +----+
              |   |     /
              |   |    /
              |   |   /
            +---------+
            | HAProxy |
            +---------+
               /  
           +----+ +----+
           | R1 | | R2 |           Redis servers
           +----+ +----+

    The scenario is simple:
      * 4 web application servers need to store and retrieve data to/from a Redis database
      * one (better using 2) HAProxy servers which load-balance redis connections
      * 2 (at least) redis servers in an active/standby mode with replication

    Configuration

    Below, is the HAProxy configuration for the

    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    12
    13
    14
    15
    16
    17
    18
    19
    20
    defaults REDIS
     mode tcp
     timeout connect  4s
     timeout server  30s
     timeout client  30s
     
    frontend ft_redis
     bind 10.0.0.1:6379 name redis
     default_backend bk_redis
     
    backend bk_redis
     option tcp-check
     tcp-check send PING
     tcp-check expect string +PONG
     tcp-check send info replication
     tcp-check expect string role:master
     tcp-check send QUIT
     tcp-check expect string +OK
     server R1 10.0.0.11:6379 check inter 1s
     server R2 10.0.0.12:6379 check inter 1s

    The HAProxy health check sequence above allows to consider the Redis master server as UP in the farm and redirect connections to it.
    When the Redis master server fails, the remaining nodes elect a new one. HAProxy will detect it thanks to its health check sequence.

    It does not require third party tools and make fail over transparent.

  • 相关阅读:
    树莓派入门之装系统
    git服务器的建立——Git折腾小记
    sk_buff的数据预留和对齐
    Linux 内核网络协议栈 ------sk_buff 结构体 以及 完全解释 (2.6.16)
    Linux网络之设备接口层:发送数据包流程dev_queue_xmit
    打印mac地址
    查看Linux进程CPU过高具体的线程堆栈(不中断程序)
    linux 终端报错 Out of memory: Kill process[PID] [process name] score问题分析
    记linux下rm误删bin文件的解决方式
    服务器有新消息主动推送给客户端浏览器
  • 原文地址:https://www.cnblogs.com/davidwang456/p/3532527.html
Copyright © 2011-2022 走看看