What steps should I follow to upgrade my VMWare Virtual Appliances Clustered PAM 2.8.3 installation to 2.8.4?

Document ID : KB000095481
Last Modified Date : 11/05/2018
Show Technical Document Details
Planning the upgrade steps procedure of a PAM installation allow us to anticipate (and minimize) the potential lack of service time an upgrade procedure requires.
What steps should I follow to upgrade my VMWare Virtual Appliances Multisite Clustered PAM 2.8.3 installation to 2.8.4?
PAM server 2.8.3 in multisite cluster with the following topology:
1 primary site with 2 or more nodes.
Several secondary sites with 2 or more nodes each.
All of them are virtual appliances.
The following procedure shows how to proceed to perform the upgrade of this clustered installation trying to have it "out of service" the minimum required time. 
There will be lack of service for some time, as the cluster has to be 'turned off' before proceeding with the upgrade. So, the upgrade should have to be a scheduled process. 

* Download the following patch "CAPAM_2.8.4.1.p.zip" from the following location: 
* Just upload it to all the PAM servers but DO NOT upgrade them by now. This allows us to save time as it can be done during production time without any impact. 
* Stop all the PAM servers. 
* When all of them have stopped, perform a snapshot of each. This will allow rolling back the upgrade in the minimum time in case of failure, returning the whole installation consistently to the point in time just before the upgrade has been started. 
* Start all the PAM servers again and make sure the cluster is properly running for all its member PAM servers, either the ones in the primary site or the ones in the secondary sites. 
* "Turn off" the cluster from the PAM client user interface. 
* Upgrade the PAM servers one by one using the following patch "CAPAM_2.8.4.1.p.zip" we previously downloaded and which is already uploaded to the PAM servers as per the two first steps of this procedure. 
* Reboot each PAM server. 
* When all the server have been rebooted, 'Turn on' the cluster again and verify the cluster is in a healthy status.