Tomcat clustering in Google compute engine for single and multiple compute engine instance

Question

I am able to configure Network Load Balancing in Google compute engine .According to my understanding and result of observation

If one tomcat configured in vm instance is down then another one will serve the request and vice-versa . Up to here i am fine. But, what if i want to i mean share session information between different vm instances .I mean to say if i want to achieve tomcat clustering within

single vm instance
Between multiple vm instance

Do Network Load Balancing also take care of sharing session information between different vm instances?

Is there infrastructure already set up for this in Google compute engine ? Or i have to manage it myself using mod_jk. .Do using teracotta server can help me on this scenario? Thank You

Answer 1

1. I have used Apache (2.2) httpd/web server;
   
2. Used mod_jk (downloaded tomcat's connector) and configured in Apache's httpd.conf and defined worker.properies
   
3. worker.properties - defined nodes (tomcat servers - i have configured for two tomcat servers) with host name, ajk port and type.
   
4. Used in-memory replication (sticky session, simpletcpcluster which uses delta manager)
   

This I have configured for clustering of tomcat servers in cloud environmnet