Repository navigation
Description
Implementation status: REST instrumentation is in #395; lifecycle-domain metrics/Events remain assigned to draft #397. Workload measurements are not complete.
Delivery bundle: A: Foundation. See #395 and #368 for the four-PR stack.
Parent: #393
Roadmap: #368
Plan item: PR-07
Priority: P2
Dependencies: None
Kubernetes support: Rolling upstream-maintained minors; use the oldest/newest boundaries pinned in config/kubernetes-versions.env. See #368 for the current policy. Unrun jobs are not passed validation.
Problem Statement
Evidence: main.go does not opt into the additional v0.25 REST-client metrics. Core controller/resource code does not register operator-specific collectors or emit transition Events.
Proposed Solution
Files: main.go, existing logging/resource/controller call sites, existing monitoring documentation/chart integration, and a small dedicated metrics source/test file if no suitable package exists.
Interface contract: Use controller-runtime's registry and RegisterRESTClientMetrics for request latency, retry, size, and rate-limiter latency as justified. Keep labels bounded; never label by task UUID or arbitrary error text.
- Add collector tests asserting the selected REST metric families are registered and representative operator counters/gauges change correctly.
- Add transition Event tests for significant lifecycle failures/progress, not one Event per poll.
- Define operator metrics for pending finalization, external task outcomes/progress, and write conflicts without replacing existing controller-runtime metrics.
- Document dashboard queries and alert/runbook meaning; baseline queue delay, reconcile duration, API requests/bytes, RSS, and startup/failover time on a reproducible workload.
- Validate package tests plus a scrape test. Record workload cardinalities and results; do not invent performance targets or claimed savings.
Done when: Subsequent optimization PRs have repeatable before/after measurements and useful failure signals.
Delivery: Four consolidated PRs, approved November 16, 2026: A -> B -> C; D branches from A. Foundation is #395. The 22 PR-XX identifiers are work-item IDs, not separate PRs. Preserve every work-item acceptance criterion and native dependency. No gated experiment is automatically enabled, and no issue is complete merely because one PR references it. D must satisfy its native #375 gate, including lifecycle-domain signals assigned to B; branching from A does not waive that merge prerequisite. Write focused regression tests first; record exact commands/results and run make check plus this issue's real-cluster/measurement gates.
Alternatives Considered
Twenty-two individual PRs add unnecessary integration overhead, while one omnibus PR obscures correctness and rollout boundaries. Deliver this scope in its approved dependency-aligned bundle. Do not expand into deferred consistency, storage, or concurrency changes without their own decision gates.
Additional Context
Baseline: 95138e0d; detailed roadmap scope: #368, section PR-07. The roadmap #368 and its native sub-issues are the planning source of truth. No planning document is committed in the final PR diffs. Re-read affected source and overlapping work before implementation. Historical baseline test passes are not validation of this change.
Activity
feat(operator): managed Kubernetes support and REST observability (#395)
amuraru commentedon Nov 17, 2026
REST instrumentation shipped in #395. Under #368, remaining measurements are scoped to #393 and the mandatory domain-metrics/Events/dashboard framework is removed. Closing the superseded broad scope, not claiming every original criterion was delivered.