How to compare and understand the Kubernetes deployment model with Pre 8.2
I am new to containerization and kubernetes. I have the following basic questions. If someone can clarify, that will be helpful for myself and many others. As I am talking to the many in the community, not many have a clear view.
Assumption: Version - 8.4.1, Server - Tomcat, Database - Postgres, Cloud - Azure
1. Can I compare JVM or node in the pre 8.2 to a POD?
2. When I want to deploy a RAP, can I just use prpcUtils or should I create a new docker image?
3. Can all PODs talk to each other? Assume I deploy a RAP into one of the PODs, does system pulse updates the every other POD and its cache?
4. When I want to horizontally auto scale a batch POD where my advanced agent is running which I want to run only in one POD (to avoid object contention), how that can be achieved without manual intervention?
5. Where do we specify JVM configuration for k8s nodes?
6. What normally differs in the CI/CD pipeline set-up when it involves k8s (AKS)
Most of the documentation I can see are for PEGA installation but not for deploying a RAP. If someone can help answering the above, it will be helpful.
k8s deployment is a new way of deploying software, which normally requires docker as container runtime. You should containerize prpcUtils into a docker image. The rest should just work. It will take time to enable k8s - including your company k8s admins, please start with this: https://github.com/pegasystems/pega-helm-charts. There are several step-by-step docs there. Also consider tiered approach with our examples of medium/large samples (clone the github repo above and check values.yaml and values-large.yaml under charts/pega).