8.7 Embedded Kafka Deprecation

In 8.7 release notes, it is mentioned like “Starting in Pega Platform 8.7, the use of the internal Kafka cluster as the Stream service is deprecated. On-premises systems that have been updated from earlier versions of Pega Platform can continue to use Kafka in embedded mode. However, to ensure future compatibility, do not create any new environments using embedded Kafka

Following are the queries that need clarification:

  • What exactly On-premisis systems mean? Are client managed clouds come under this?
  • While updating from 8.4 to 8.7 in client managed cloud/private cloud, Is embedded kafka available in 8.7 after updation? (or) Integrating with external kafka is mandatory?

Any update on this?

Hi Pega Team,

Recently we are in between upgrade from Pega 8.3.1 to 8.7.1 , I have as well follow up for same questions

Thanks

Sujeet

@AjitChandraM9485 @SPSUJITMCA this statement applies for both on-premises and managed/private cloud deployments. Upgrading customers, like the one you’ve described in your question, can continue to use the embedded Kafka. However, we highly recommend starting planning the transition to externally managed Kafka.

For managed/private cloud deployment, we are about to release a new version of our helm charts that will make it easy to connect to an external Kafka cluster. You can follow new releases here: https://github.com/pegasystems/pega-helm-charts/releases

@adama , Hi Andrei, is this statement also valid for in place upgrade from v8.6.1 to v8.8.x ? We are moving from v8.6.1 to v8.8.4 on Redhat openshift container platform.

@AnkurG0647 as already answered in the duplicate post Is dedicated Kafka cluster a must? where you also posted your question, here is the answer:

"Usage of Embedded kafka has been deprecated in Pega v8.7.x & Pega v8.8.x. It’s advised to use an external Kafka since the support will be removed in future pega versions.

To answer your query, all existing older pega deployments should able to upgrade (till pega v8.8.x) with embedded kafka as option".

@MarijeSchillern , awesome thanks heaps for this swift response :slight_smile: