Deploying Pega to K8s with Postgres Database
Hi, we are installing Pega 8.3 on kubernetes platform with a Postgres backend database. We have a few questions that we would like to be answered.
1. For Postgres installaion, it requires PL/Java module installed on the database server. Our company guidance does not allow us to have jvm on database side. So we have the following questions brought up by our architecture team:
- Why do we need JVM on database side
- Does Pega have any future roadmap to remove it?
- Are there any other alternatives to having to install JVM on DB side?
- What's the footprint of the JVM, memory usage, etc?
2. To install Pega image on Kubernetes, it needs certain permissions. We had permission denied issue (when the deployment is trying to chmod on a folder). We would like to know what kind of user permission the deployment needs to have so our k8s team to configure it.
Errors we got:
- If we don't specify runAs, it will give us "Error: container has runAsNonRoot and image will run as root";
- If we specify runAs: 65534, it gives us "standard_init_linux.go:207: exec user process caused "permission denied"".
TIA.
Capgemini
FR
Hi,
No way to avoid using java in database, Pega uses it a lot in SQL procedures.
For the installation, you need to create specific images. Do not run Pega installation script inside a tomcat image ! You need to create you own container images with Pega jar extracted in it. Or use Pega managed docker image (https://hub.docker.com/r/pegasystems/pega)
Regards,
Marc-Antoine
Pegasystems Inc.
US
Shenyu,
Your first question is related to UDF, it was meant only for dev/test environment and should never be used in a production environment (you can byapss UDF deployment during rule base loading). Regarding the roadmap, you should engage your Pega account manager to connect with Pega product management. One UDF alternative is plv8, which is a javascript plugin for postgres, that is what is deployed in pegacloud (as AWS does not allow pljava either). But again, much slower than pljava and can only be used in dev/test.
As for the second, I wonder if you can provide a bit more details in your environment. What I am wondering is that I have never seen that deploying in my internal open source k8s cluster (hosted on centOS), using OOTB helm charts.