Allow StartupProbe configuration for the deployment (via values.yaml)

[karawedi]

Is your feature request related to a problem? Please describe.

Running Node Red with many extensions can heavily impact startup time. I'd like to be able to implement agile liveness/readiness Probes without having to pray node-red starts all flows before the implemented initialDelaySeconds runs out and the container is being killed.

Running higher initialDelaySeconds would mean loosing sensitivity on the Probes actually measuring the flows responsiveness. I therefore suggest, the K8s default mechanics should be used - I'm aware of the implications (especially) on the sidecar and will try to document this as good as possible.

Describe the solution you'd like

The startupProbe should be made available - the same as the liveness and readiness Probe are implemented already.

Describe alternatives you've considered

increasing initialDelaySeconds on Probes: This works, but (imo) works against making the service available as fast as possible after startup, while still being able to handle the long startup of Node-Red. The StartupProbe is the response of Kubernetes to this exact problem.

Search

• I did search for other open and closed issues before opening this.

Code of Conduct

• I agree to follow this project's Code of Conduct

Additional context

No response

[karawedi]

Hi everybody! I'll implement this change and will create a PR once finished and tested on my end. Feel free to get in touch, if I'm overlooking something - especially if you actively decided against implementing the startupProbe some time ago.

I also saw the open PR by @christf, which I'd maybe contribute to, especially concerning docs.