Running Prisma Cloud Compute in the clustered-DB configuration lets you sync Console’s database across multiple instances of Console. When using the clustered-DB deployment, all Consoles must be accessible with the same DNS name and connected to the same load balancer.
The Consoles in a clustered-DB pool can be deployed in the same region across different availability zones.
A new Prisma Cloud Compute self-hosted setup is required to create a clustered-DB deployment. That is, all Consoles must be new. Accordingly, all Defenders must be new as well.
Recommendations for your clustered-DB setup
The information in the following sections help you make informed decisions when selecting pool size, load balancer configuration, and storage types.
Number Of Consoles in the clustered-DB pool
The clustered-DB mechanism continuously requires the pool to choose a primary Console. This is done by the replica sets election mechanism. The election determines which member will become primary. Replica sets can trigger an election in response to a variety of events, such as: adding a new member, initiating a pool, and losing connectivity to the primary for more than the configured timeout.
An odd number of Consoles is recommended for better fault tolerance. The table below describes the fault tolerance for 3/5/7 Consoles:
Number of Members
Majority Required to Elect a New Primary