We have implemented Pega HA into our environment, but are seeing mixed results. In some cases when we initially perform a fail over test by stopping Apache Tomcat or rebooting a WebUser node, the user will always get redirected to login to another node, but may not get a message that a crash event occurred, the UI is not always recreated, and data is never committed unless it was already saved by hitting the save button. We have checked Dynamic System Settings to ensure "enable server crash recovery" and "enable end user messaging of crash event" are enabled. And did a rolling reboot of each server in the cluster. This system is also using the default database storage method.
Any help would be appreciated.
***Edited by Moderator Marissa to update platform capability tags****
It definitely does not include complete settings (e.g., session/ha/enabled), also you need to make sure several settings are of type prconfig, which means if you want to use DSS, it should follow prconfig/<setting>/default format. Finally, what load balancer are you using in front of tomcat? Follow this guide https://community.pega.com/system/files/pdfs/Pega%207.4%20High%20Availa…
I know this is 7.4, but should be accurate for basic setup.