I've been preparing for HA option on Pega 7.4 and found a possible contradiction in the Pega 7.4 High Availability Administration Guide. The description of setting session/ha/Enabled differs on page 10 and 18.
The page 10 says in "Enabling high availability" says:
To enable high availability on a Pega Platform node, set the session/ha/Enabled parameter to true in the prconfig.xml file. For example:
<env name="session/ha/Enabled" value="true"/>
Whereas page 18 says that
You can display or hide high availability landing pages by adjusting the following prconfig setting.
• Value Type — Boolean
• Value — true or false
• Functionality — Switch to turn on or off the landing page user interface (UI)
Furthermore, it says that
Landing pages ... are not recommended due to latency of system pulse communication
What is the proper value of session/ha/Enabled if I want HA option but try to follow the recommendation and don't want to use landing pages?
The setting "session/ha/Enabled" happens to be dual purpose. It both enables and disables High Availability from running and is also used to allow access to the High Availability landing pages. In addition to this setting, for a user to have the privileges needed to access the HA landing pages, their access group must also have the "PegaRULES:HighAvailabilityAdministrator" role.
As for the bullet regarding "Landing pages ... are not recommended due to latency of system pulse communication", this is a warning from the 7.1.x series of the product (that I'll need to put in a request to remove!). The System Pulse used to operate through the database and required up to two minutes for changes to propagate through the system. As of 7.1.9, the System Pulse is operated through our clustering services and happens near instantaneously. That comment was to steer customers away from using these landing pages since they might wait a while for the operations to take effect. Pega 7.4 uses the cluster-based System Pulse and requests will be processed immediately.