feat(health): C3b — internal/health thin readiness poller (spec 10) - #4
Merged
Conversation
New read-only, lock-free package that gates the up saga on real readiness: - Compile(*config.Healthcheck) → Timing with spec-10 defaults (interval 5s, timeout 3s, retries 10, startPeriod 30s); the "60s for stateful images" rule stays in the engine templates (no hardcoded engine knowledge here). Budget() = startPeriod + interval*retries. - Poll(ctx, docker.Client, Target, Timing): polls .State.Health.Status until Healthy (or Started: container running), failing fast on unhealthy / a fatal container state (exited/dead/restarting) / timeout / ctx cancel. Transient inspect errors (container not created yet) are tolerated until the deadline. - On failure returns a *ProbeError carrying the --json Record AND the last 20 log lines (fetched on a detached context so a timed-out poll still has a diagnostic) + a one-line remediation (ARCHITECTURE §7.6). - Record is the machine-readable health contract (service/project/kind/status/ attempts/elapsedMs/lastError, lastError null on success). Thin v1 — single-target Poll consumed linearly by the caller (C5). The full workspace DAG (cycle paths, topo waves, profile pruning, generate-time "condition:healthy needs a healthcheck") is X2. Unit tests use a scripted client (healthy-after-starting, fail-fast unhealthy/ exited, started condition, timeout, transient-error recovery, ctx cancel, Compile defaults/overrides/bad-durations). A `//go:build integration` test polls real healthy + unhealthy containers and asserts the log diagnostic. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
gustavobertoi
added a commit
that referenced
this pull request
Jun 29, 2026
…& packages (#77) Completes the autonomously-doable part of G4 (release plumbing). goreleaser's default archive globs bundle LICENSE*/README* but NOT NOTICE — an Apache-2.0 distribution must ship NOTICE. Adds an explicit archive `files:` list (LICENSE + NOTICE + README + docs/QUICKSTART) and embeds LICENSE+NOTICE under /usr/share/doc/devstack in the .deb/.rpm via nfpm `contents:`. Validated locally: `goreleaser check` passes and a full `--snapshot` build produces all 8 artifacts (4 tar.gz + 2 .deb + 2 .rpm) with NOTICE present in the tarball. release-dryrun continues to cover this on every PR. Remaining G4 is owner-only (outward-facing / irreversible, locked decision #4): create the Homebrew tap repo (Q-NAME) + wire `brews:`, then cut the release — `git tag v1.0.0 && git push origin v1.0.0` — and flip the repo public. Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Chunk C3b (M2-remainder)
New read-only, lock-free
internal/healthpackage — the readiness gate the up saga (C5) stands on. Per spec 10 (thin v1).What
Compile(*config.Healthcheck) → Timingwith spec-10 defaults (interval 5s, timeout 3s, retries 10, startPeriod 30s). The "60s for stateful images" rule lives in the engine templates, not here (no hardcoded engine knowledge).Budget() = startPeriod + interval*retries.Poll(ctx, docker.Client, Target, Timing)— polls.State.Health.StatusuntilHealthy(orStarted: container running), failing fast onunhealthy/ a fatal container state (exited/dead/restarting) / timeout / ctx cancel. Transient inspect errors (container not created yet) are tolerated until the deadline.*ProbeErrorcarrying the--jsonRecordand the last 20 log lines (fetched on a detached context so a timed-out poll still has a diagnostic) + a one-line remediation (ARCHITECTURE §7.6).Recordis the machine-readable health contract (service/project/kind/status/attempts/elapsedMs/lastError,lastErrornull on success).Scope
Thin v1 — single-target
Pollconsumed linearly by the caller. The full workspace DAG (cycle paths, topo waves, profile pruning, generate-time "condition:healthy needs a healthcheck") is X2.Tests
//go:build integrationtest polls real healthy + unhealthy containers and asserts the log diagnostic (verified locally against Engine 29.5.3).Gate
make cigreen;go test -tags=integration ./internal/healthgreen locally.🤖 Generated with Claude Code