You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@geronimo.apache.org by "Ivan (JIRA)" <ji...@apache.org> on 2009/11/19 15:14:39 UTC
[jira] Resolved: (GERONIMO-4955) Tomcat Clustering via Uni Cast
doesn't work
[ https://issues.apache.org/jira/browse/GERONIMO-4955?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Ivan resolved GERONIMO-4955.
----------------------------
Resolution: Fixed
Fix Version/s: 2.1.5
For 2.1.4, it should be fixed after merging the patch of Geronimo-4187 to 2.1 branch.
> Tomcat Clustering via Uni Cast doesn't work
> -------------------------------------------
>
> Key: GERONIMO-4955
> URL: https://issues.apache.org/jira/browse/GERONIMO-4955
> Project: Geronimo
> Issue Type: Bug
> Security Level: public(Regular issues)
> Components: Clustering
> Affects Versions: 2.1.4, 2.2
> Environment: Server 1:
> OS:Windows Vista amd64 6.0 build 6001 Service Pack 1
> Geronimo: 2.2-SNAPSHOT Build 2009.09.28-12:26:20.913+0800
> JVM: JDK 6
> Server 2:
> OS: Windows Server 2008 amd64 Version 6.0 build 6002 Service Pack 2
> Geronimo:2.2-SNAPSHOT 2009.09.28-12:26:20.913+0800
> JVM:JDK 6
> Reporter: Lu Jiang
> Fix For: 2.1.5
>
> Attachments: node02ExcerptFromConfig.xml.txt, node03ExcerptFromConfig.xml.txt, realm.xml, servlet-examples-cluster-plan2.xml, servlet-examples-cluster-plan3.xml, servlet-examples-cluster-server2.war, servlet-examples-cluster-server3.war, simple-servlet-examples-cluster-plan2.xml
>
>
> Below is a senario:
> 1.Install geronimo 2.2 on two machines,namely 9.186.10.60(NODE02),9.186.10.89(NODE03).
> 2.Configure Apache http server.
> (1)add mod_jk.so to ApacheHttpServer_HOME/modules.
> (2)add or modify workers.properties under ApacheHttpServer_HOME/conf like
> worker.list=loadbalancer,status
> worker.maintain=60
> worker.NODE02.port=8009
> worker.NODE02.host=9.186.10.60
> worker.NODE02.type=ajp13
> worker.NODE02.lbfactor=1
> worker.NODE03.port=8009
> worker.NODE03.host=9.186.10.89
> worker.NODE03.type=ajp13
> worker.NODE03.lbfactor=1
> worker.loadbalancer.type=lb
> worker.loadbalancer.balance_workers=NODE02,NODE03
> worker.loadbalancer.sticky_session=1
> worker.status.type=status
> (3)add information below to ApacheHttpServer_HOME/conf/
> LoadModule jk_module modules/mod_jk.so
> JkWorkersFile conf/workers.properties
> JkLogFile logs/mod_jk.log
> JkLogLevel debug
> JkMount /* loadbalancer
> 3.
> For NODE02:
> (1)Modify GERONIMO_HOME/var/config/config-substitutions.properties
> ClusterNodeName=NODE02
> (2)Modify GERONIMO_HOME/var/catalina/server.xml.
> <Engine name="Catalina" defaultHost="${ServerHostname}" jvmRoute="NODE02" >
>
> For NODE03:
> (1)Modify GERONIMO_HOME/var/config/config-substitutions.properties
> ClusterNodeName=NODE03
> (2)Modify GERONIMO_HOME/var/config/config.xml.
> <Engine name="Catalina" defaultHost="${ServerHostname}" jvmRoute="NODE03" >
>
> 3.
> I tried two ways to enable unicast clustering..One is to enable it in config.xml and the other is to enable it in the deploy plan.
> (Deploy the ream.xml first cause we have security realm configuration in the application )
> The first way:(enable it in config.xml)
> On the two server,separately add GBean information under<module name="org.apache.geronimo.configs/tomcat6/2.2-SNAPSHOT/car">
> Refer to the attachment:node02ExcerptFromConfig.xml and node03ExcertFromConfig.xml
> Then deploy the application servlet-examples-cluster-server2.war using Simple simple-servlet-examples-cluster-plan2.xml as a deploy plan.
> deploy the application servlet-examples-cluster-server3.war using Simple simple-servlet-examples-cluster-plan3.xml as a deploy plan.
> The second way:(enable it in deploy plan)
> Without modification to config.xml,deploy the application directly.
> deploy the application servlet-examples-cluster-server2.war using servlet-examples-cluster-plan2.xml as a deploy plan.
> deploy the application servlet-examples-cluster-server3.war using servlet-examples-cluster-plan3.xml as a deploy plan.
> Expected:
> Session will be replicated .When node 02 fails,the node 03 will be the back up.and vice versa.
> But currently both way don't work, when node2 fails ,session is not replicated to node03
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.