Better handling of PATH env var with initConfig

[vmdude]

what

We got a use-case where we need to defined the PATH environment variable through the .Values.environment node for the statefulset, and we had some issues due to duplicate entries.
When .Values.initConfig is enabled, the chart actually add a PATH environment variable.
The issue is due to how Kubernetes handle duplicate envvars definition, when a referenced key is present in multiple resources, the value associated with the last source will override all previous values.
As .Values.initConfig variable is defined later in the manifest, it will always supersede previously definition, thus our .Values.environment node was bypassed.

This PR will add the following logic:

• if .Values.initConfig.enabled and no PATH in .Values.environment: add a PATH envvar like before
• if .Values.initConfig.enabled and PATH in .Values.environment: do not add a PATH envvar as it should use the one defined in .Values.environment

The rest of the logic is untouched.

why

We wanted to be able to define more specifically how the PATH environment variable is presented to the statefulset avoiding any duplicate in the process.

tests

• I have tested my changes by created several values file for all possible combination regarding PATH and initConfig parameters

references

• More information about precedence and duplicate env var management [so]

[GMartinez-Sisti]

Spot on 🚀