Query about resource limits and requests

[mandarbhosle]

For namespaces having resource quotas, the netcheck assertion pod doesn't come up. As it doesn't have the resource limit and request.

I tried adding it to the values.yaml file under probeconfig and did reinstall. Still it doesn't pick up.

Is it something i am missing. Also in the configmap.yaml templa didn't find the code to pick it up.

Also is there any parameter to reduce the wait time for fail check rules

[hardbyte]

That's certainly a valid feature request. The way to support it would be to modify the configmap.yaml in the helm template to persist all of probeConfig to the configmap.

Adjust the create_job_spec in operator/netchecks_operator/main.py to use the given resources similar to how the image and pull policy is loaded from the config:

container = client.V1Container(
        name="netcheck",
        # e.g "ghcr.io/hardbyte/netchecks:main"
        image=f"{settings.probe.image.repository}:{settings.probe.image.tag}",
        image_pull_policy=settings.probe.image.pullPolicy,

settings.probe is defined as a Pydantic BaseModel in netchecks_operator/config.py - that actually already has resources.

Happy to review a PR if you'd like to implement it.

[mandarbhosle]

Hello,

Pull request https://github.com/hardbyte/netchecks/pull/155 to include the resource request and limits. Request your review

Also is there any parameter to reduce the wait time for fail check rules. Currently it takes 2 min to report a failure rule. So it takes time to run the entire policy if there are multiple rules.

[mandarbhosle]

Hello,

Updated the code further to include resource quotas, node affinity

Regards, Mandar Bhosle