Support for embedded deployment mode for Pega 24.2
Starting with Pega Platform 24.2, externalization of key services such as Hazelcast, Elasticsearch (Search), and Kafka is required. Due to certain dependencies, our current approach is to initially run Pega 24.2 on a VM-based architecture, and then transition to a container-based (Kubernetes) architecture in the coming months.
According to the release notes:
-
Embedded Hazelcast is still supported in Pega 24.2, but it is deprecated and will be completely unsupported starting with Pega Platform 25.
However, we would like to understand the current support status for Search and Kafka in Pega 24.2:
-
Can Search (Elasticsearch) still be run in embedded mode, or is externalization mandatory?
-
Is Kafka supported in embedded mode, or does it also require externalization from Pega 24.2 onward?
We would appreciate any insights or guidance on how these services must be configured during this interim VM deployment phase.
HCA Healthcare
US
@NK Bitta
In Pega 24.2, you must use external systems for both Search (Elasticsearch) and Kafka, as embedded mode is no longer supported for them. Only Hazelcast can still run in embedded mode for now, but it is deprecated and will be removed in Pega 25, so it's better to plan for external Hazelcast soon. For your current VM-based setup, you can keep using embedded Hazelcast temporarily, but Search and Kafka must be set up externally. This setup ensures your system works properly and is ready for future upgrades. When you move to containers later, having these services already externalized will make the transition smoother