[CASCL-1304] kubectl-datadog: enrich dd-cluster-info ConfigMap

[L3n41c]

What does this PR do?

Enriches the dd-cluster-info ConfigMap (introduced by #2945) so a future migration tool can:

• Distinguish Datadog-managed node-management entities (Fargate profiles, Karpenter NodePools) from legacy ones to drain — via a per-entity managedByDatadog flag.
• See the running Karpenter installation (version, namespace, ownership) under a new autoscaling parent that also groups the existing clusterAutoscaler entry and a new eksAutoMode entry.

As a side-effect, two detection helpers (FindKarpenterInstallation, IsEKSAutoModeEnabled) move from install/guess/ to new common/karpenter/ and common/eksautomode/ packages so the clusterinfo classifier can reuse them. A generic commonk8s.FindFirstDeployment factors out the shared pager+predicate scan, and commonk8s.ExtractDeploymentVersion factors out the controller-image-tag → label fallback used by both the Karpenter and Cluster Autoscaler detectors.

Motivation

Follow-up to #2945 (CASCL-1304). The original snapshot only captured nodes grouped by their owning manager; the future migration tool also needs to tell Datadog-managed managers (to keep) from legacy ones (to drain), to know whether Karpenter is already running, and to know whether EKS auto-mode is active so it can short-circuit when there is no migration to drive.

Additional Notes

• Schema change is breaking, on purpose: no consumer of dd-cluster-info exists yet (grep -r "dd-cluster-info\|ConfigMapDataKey" --include='*.go' returns only the writer). APIVersion stays at v1.
• NodePool ownership detection uses the broader autoscaling.datadoghq.com/created label alone — not uninstall.go's AND-pair with app.kubernetes.io/managed-by: kubectl-datadog. The cluster agent creates NodePools with only the created label, and the migration tool must preserve them too. This divergence is deliberate.
• Fargate profile ownership detection reads tags via EKS.DescribeFargateProfile. The expected managed-by: kubectl-datadog tag is propagated automatically from the CloudFormation stack tags written by common/aws/cloudformation.go, so no infrastructure change is needed.
• Datadog-managed NodePools with no nodes yet (typical right after install) are seeded into the snapshot with an empty nodes list so the migration tool sees the destination NodePools exist.
• Best-effort detection: every new external API call (EKS DescribeFargateProfile, NodePool list, Discovery for auto-mode) is tolerated — transient errors / missing CRDs log a warning and leave entries unflagged rather than failing the snapshot. The call site (recordClusterInfo) was already best-effort before this PR.

Minimum Agent Versions

• Agent: N/A (kubectl-datadog plugin only)
• Cluster Agent: N/A (kubectl-datadog plugin only)

Describe your test plan

Automated coverage added in this PR:

• TestClassify_KarpenterNodePoolOwnership: kubectl-datadog (both labels) + cluster agent (created label only) + Datadog NodePool with no nodes yet + foreign NodePool.
• TestClassify_KarpenterNodePoolOwnership_NoCRD: tolerant of meta.IsNoMatchError when the Karpenter CRD is not installed.
• TestClassify_FargateProfileOwnership / TestClassify_FargateProfileOwnership_DescribeError: tag-based detection + AWS API error fallback.
• TestClassify_KarpenterDetection: version extraction from controller image tag, ManagedByDatadog/InstallerVersion from sentinel labels.
• TestClassify_EKSAutoMode: discovery API exposes nodeclasses → Enabled: true.
• TestPersist_YAMLShape: pins lowerCamelCase wire keys against the gopkg.in/yaml.v3 lower-case-by-default footgun.
• TestFindFirstDeployment_*: covers the new generic helper.

Manual test plan on a sandbox EKS cluster:

• kubectl datadog autoscaling cluster install succeeds.
• kubectl get cm -n dd-karpenter dd-cluster-info -o yaml contains:
  • autoscaling.karpenter.{present, version, managedByDatadog, installerVersion}
  • autoscaling.eksAutoMode.enabled
  • nodeManagement.fargate."dd-karpenter-<cluster>".managedByDatadog: true
• At least one NodePool created by kubectl-datadog appears under nodeManagement.karpenter with managedByDatadog: true, even with no node landed on it yet.
• A hand-created foreign NodePool (no Datadog labels) does NOT carry managedByDatadog: true.

Checklist

• PR has at least one valid label: enhancement, refactoring
• PR has a milestone or the qa/skip-qa label
• All commits are signed

[codecov-commenter]

Codecov Report

❌ Patch coverage is 72.15190% with 44 lines in your changes missing coverage. Please review.
✅ Project coverage is 41.42%. Comparing base (d1d2b65) to head (992bd28).
⚠️ Report is 2 commits behind head on main.

Files with missing lines	Patch %	Lines
...tadog/autoscaling/cluster/common/k8s/deployment.go	45.45%	18 Missing ⚠️
...autoscaling/cluster/common/clusterinfo/classify.go	84.94%	10 Missing and 4 partials ⚠️
...bectl-datadog/autoscaling/cluster/install/steps.go	20.00%	12 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2980      +/-   ##
==========================================
+ Coverage   41.39%   41.42%   +0.02%     
==========================================
  Files         331      332       +1     
  Lines       28911    28984      +73     
==========================================
+ Hits        11969    12007      +38     
- Misses      16086    16118      +32     
- Partials      856      859       +3     

Flag	Coverage Δ
unittests	41.42% <72.15%> (+0.02%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
...oscaling/cluster/common/eksautomode/eksautomode.go	88.88% <100.00%> (ø)
.../autoscaling/cluster/common/karpenter/karpenter.go	92.85% <100.00%> (ø)
...bectl-datadog/autoscaling/cluster/install/steps.go	19.51% <20.00%> (-0.49%)	⬇️
...autoscaling/cluster/common/clusterinfo/classify.go	86.95% <84.94%> (-3.96%)	⬇️
...tadog/autoscaling/cluster/common/k8s/deployment.go	45.45% <45.45%> (ø)

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update d1d2b65...992bd28. Read the comment docs.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.