Apache Tomcat Connection and
Failover-Cluster
For
high availability of web applications based on OpenCMS Platform at
Pomegranate Portal we provide a Cluster of two Tomcat Servlet Engines,
which are hosted on different machines. The Apache Web server, which
resides on a separate machine, has the role to forward the client
requests to the tomcat servers with load balancing features. Figure 1
illustrates a single environment without clustering and shows how
Apache and Tomcat could be connected together via mod_jk module over
the AJP (Apache Java Protocol)Connector. Mod_jk works by configuring
Apache to load the mod_jk module. When Apache starts, it creates a
series of connections with Tomcat over the JK connector, usually on
port 8009. A short HowTo could be found on our forum.
Figure 1: Connecting Apache to
tomcat with mod_jkThis environment can be extended to
a load balanced cluster environment. A cluster is a group of load balanced
Tomcat instances, which have the same session data. The sessions will be
replicated between Tomcat instances via the DeltaManager of
each instance. Clustering works via multicast addressing, whereby all Tomcat
instances broadcast their session information on the same port, but listen on
different ports. The following Figure 2 shows the cluster architecture with 2
load balanced Tomcat instances.
Figure 2: Appache - Tomcat Cluster Architecture.
If one instance
fails or is stopped, due to maintenance or Soft- or Hardware failure, the other
instance will pick up the session and have all data available and Apache will
be noticed that one Tomcat instance is out of state and redirects the client
requests to the next member in the cluster.
For setting up
such an environment we have several options:
One Apache Web
server on its own machine and two tomcat servers on different machines (stable
and high performance, high cost)
One Apache Web
server on its own machine and two tomcat instances on the same machine.
One Apache Web
server and one tomcat server on one machine and the 2. Tomcat server on a separate
machine.
分享到:
相关推荐
分布式NOSQL redis源码阅读中文分析注释,带详尽注释以及相关流程调用注释,提出改造点,redis cluster集群功能、节点扩容、槽位迁移、failover故障切换、一致性选举完整分析,对理解redis源码很有帮助,解决了...
Flume NG 是Cloudera提供的分布式数据收集系统,它能够将不同数据源的海量日志数据进行高效的收集、聚合、移动,最后存储到存储中心。Flume NG支持(故障转移)failover和负载均衡。
Failover Cluster Step-by-Step Guide - Configuring a Two-Node File Server Failover Cluster
Flume 是Cloudera提供的分布式数据收集系统,它能够将不同数据源的海量日志数据进行高效的收集、聚合、移动,最后存储到存储中心。Flume NG支持(故障转移)failover和负载均衡。
Flume 是Cloudera提供的分布式数据收集系统,它能够将不同数据源的海量日志数据进行高效的收集、聚合、移动,最后存储到存储中心。Flume NG支持(故障转移)failover和负载均衡。
Redis 连接管理器, 防灾难, 实现故障重连.
simple-failover-java A simple failover library for Java. 用于构建高性能的客户端(主调方)自适应负载均衡和自动重试能力。 jdk1.8+ Get Started // 添加多个被调用资源,这里的被调用资源是指目标服务器(有多...
2008下,cluster的配置。需要的用户可以根据此文档,正确的配置cluster。
NULL 博文链接:https://acooly.iteye.com/blog/1347417
Failover Clustering adds convenience and protections to keep those systems running. A well-designed cluster requires understanding many topics, but the time invested in learning pays off with a ...
一共9页,简单但是很清晰,比几十页看完让人晕的什么也不会了强多了! 不知道现在还有没有人用2008,希望对需要的人有些帮助
murmur-failover-daemon通过定期将主数据库同步到从数据库来工作。 同时,不断对主机进行ping操作,以查看主机是否启动。 如果主服务器出现故障,则从服务器将启动,并开始接受连接。 与主服务器断开连接的Mumble...
气流计划程序故障转移控制器项目目的该项目的目的是创建一个故障转移控制器,该控制器将控制哪个调度程序已启动并正在运行,以允许跨整个Airflow集群进行HA。动机我们尝试设置一个高可用性气流集群,其中有两台运行...
Flume 是Cloudera提供的分布式数据收集系统,它能够将不同数据源的海量日志数据进行高效的收集、聚合、移动,最后存储到存储中心。Flume支持(故障转移)failover和负载均衡。
详细描述sqlserver2008在windows server2008 R2上实现failover cluster的每一步骤
我们可以将多台服务器组成一个 故障转移群集 (failover clustercluster),这 些服务器 会协同工作 以提供一个 高可用的 应用程序 或服务环境 。群集内每一台服 务器被称为一个 节点 (node ),节点之间通过网络...
windows r2 failover cluster whitepaper
详细描述了SQL Server2008 failover cluster 的搭建过程,包括域创建、scsi创建,server failover、sqlserver cluster,每一步都有详细的说明及截图
Yii 数据库故障转移 如果数据库连接失败,它将尝试连接到其他故障转移连接。 用法 ... 在这种情况下,如果db连接失败,它会先尝试连接db3,如果db3失败,它会尝试db4。 'db' => array ( 'class' => 'application....
实现了自动安装配置redis 已经测试过了,如果有问题,请留言 启动脚本需要 3个参数 serverIP masterIP redisType 例如, 作为master ./redis_create.sh 192.168.10.10 192.168.10.10 master 作为slave ...