feat: allow overriding instance label

[fad3t]

Allow overriding app.kubernetes.io/instance label per workload

Summary

PR #2111 (released in 3.140.0) changed the app.kubernetes.io/instance label on the Agent, Cluster Agent, and Cluster Checks Runner workloads from <release-name> to <fullname>-<workload> (e.g. datadog-agent, datadog-cluster-agent). This is a breaking change for users whose NetworkPolicies — or any other selector — pin on the previous value.

agents.additionalLabels doesn't solve the problem: it appends labels, producing duplicate keys when the key already exists in the templated label set.

This PR adds three new values:

• agents.instanceLabelOverride
• clusterAgent.instanceLabelOverride
• clusterChecksRunner.instanceLabelOverride

When set, each overrides the app.kubernetes.io/instance label on the corresponding workload (both metadata.labels and spec.template.metadata.labels). When unset (default), behavior is unchanged from current main.

Usage

Restore the pre-3.140.0 label so existing NetworkPolicies keep matching:

agents:
  instanceLabelOverride: datadog
clusterAgent:
  instanceLabelOverride: datadog
clusterChecksRunner:
  instanceLabelOverride: datadog

Why scoped to a single label

PR #2111 only changed app.kubernetes.io/instance. The other three app.kubernetes.io/* labels emitted by datadog.pod-template-labels were unaffected, so a generic labelOverrides map would be wider scope than needed. Keeping it to one string value keeps the helper diff minimal and the values surface obvious.

Safety

app.kubernetes.io/instance is not part of any DaemonSet/Deployment selector.matchLabels (selectors use app: <fullname>...), so overriding the value does not run into immutability issues on upgrade.

Test plan

• helm template with defaults: output is byte-identical to current main for the relevant manifests
• helm template … --set agents.instanceLabelOverride=datadog --set clusterAgent.instanceLabelOverride=datadog --set clusterChecksRunner.instanceLabelOverride=datadog: confirm the app.kubernetes.io/instance value flips on the agent DaemonSet, cluster-agent Deployment, and cluster-checks-runner Deployment (both metadata and pod template)
• Existing CI matrix passes
• helm-docs regenerated (README in sync)

Related

• Original change: Update agent, cluster-agent, and cluster-checks-runner pod labels #2111

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 9ca6d3345d

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".

[fad3t]

Related to #2309

[fad3t]

Let me know if there's anything else required to move this one forward :) Thanks!

[fad3t]

Can you help move this one forward please? Thanks @tedkahwaji !

[fad3t]

Any maintainer to approve the workflow please?