Table of Contents

Search

  1. Preface
  2. Introduction to High Availability
  3. Set Up the High Availability Environment
  4. Configure JMS Discovery
  5. Configure RMI Connections
  6. Configure the Operation Console
  7. Configure an HTTP Load Balancer
  8. Configure the Dashboard and Reports
  9. Appendix A: Troubleshooting High Availability
  10. Appendix B: High Availability Log Messages
  11. Appendix C: Port Number Reference
  12. Appendix D: Glossary

High Availability Guide

High Availability Guide

Configuring Sticky Sessions

Configuring Sticky Sessions

A sticky session is a session that uses a single server to handle all user traffic for a specific browser. In a sticky session, after the browser creates a session, the same server that received the first session continues to process all requests from that session until the user logs out of the browser.
When you enable sticky sessions, the load balancer forwards all requests from a specific browser to the same server until the user ends the session, regardless of the load on the servers in the cluster. For example, if browser 1 starts a session and the load balancer forwards the initial request to Tomcat A, all subsequent requests are sent to Tomcat A.
Sticky sessions do not require multicast transmission to multiple destinations and do not generate additional network traffic or reduce performance. However, if a server fails, all existing sessions on that server end without a failover mechanism and all users must manually log in to a different server.

0 COMMENTS

We’d like to hear from you!