Hello,

I'm trying to implement High Availability features on my Pega 8.6 Platform related to the following documentation : 

https://community.pega.com/system/files/media/pdf/pega-platform-81-high-availability-admin-guide.pdf

I'm facing issue with the Crash Server Recovery feature.

The passivation has been set in "Database" shared storage and the DSS prconfig/session/ha/crash/RecordWorkInProgress/default is set to "true"

All nodes in the cluster have been restarted.

1) We have 2 webuser nodes (Tomcat server) behind the load balancer

2) We are using SSO with SAML 

3) When I'm using the Quiesce method (Drain the node on LB, Quiesce the node on Pega and after complete Quiesce ,disable the node on LB) , the session is correctly redirected to the second node with the same requestor and all tabs etc... are recovered.

4) But If I crash a node (kill Tomcat JVM), the session never recovered on the second node (http redirection is OK) : we automatically have a page which ask login/pwd and Login with SSO button. We can see , in the response request that the server (second node so) send, Pega-RULES cookie with a new requestor.

How the second node can recover a session (requestor) in case of crash ?

Where we can see the data stored after RecordWorkInProgress activation, how does it work?

Thanks to help me if you allready have successfully implement this feature!