beads/release-gates/be-u8ldr-multiprocess-schema-init-gate.md
Jim Wordelman 6deca80cf2
fix(dolt): remove dead decoy-listener test code, use retry-safe schema init under 10-way contention (#5880)
* feat: green — Fix: multiprocess tests INSERT without issues.description and report OK on zero writes; dolt init has no identity (refs be-kafb)

* test(feat): red — Review: Fix: multiprocess tests INSERT without issues.description and report OK on zero writes; dolt init has no identity (refs be-yzowm)

* feat: green — Review: Fix: multiprocess tests INSERT without issues.description and report OK on zero writes; dolt init has no identity (refs be-yzowm)

* fix(dolt): remove dead decoy-listener test code, use retry-safe schema init under 10-way contention

Round-2 review (be-4oah9) found two blocking issues in
TestMultiProcessSchemaInit_DoltVerify:

1. The "hermetic regression test" decoy listener set
   BEADS_DOLT_SERVER_PORT to its own port, then immediately cleared it
   to "" on the next line before doltserver.Start ran — so the decoy
   was never actually exercised. Removed the dead block; kept the
   actual fix (clearing the ambient env var before Start) with a
   comment describing why it's needed instead of a non-functional
   simulation of it.

2. TestMultiProcessSchemaInit_DoltVerify failed reproducibly on a
   migration-lock-timeout distinct from round 1's port collision.
   Root cause: the test's 10 concurrent goroutines called the raw,
   single-attempt initSchemaOnDB instead of initSchemaOnDBWithRetry —
   the wrapper real bd processes use via store.go's initSchema
   specifically so contended GET_LOCK attempts converge instead of
   one unlucky waiter failing outright. Sibling file
   initschema_concurrent_test.go already uses the retry variant for
   the same kind of contention scenario.

Verified: TestMultiProcessSchemaInit_DoltVerify PASS x3 (incl. once
with ambient BEADS_DOLT_SERVER_PORT set, matching the reviewer's
environment), plus TestMultiProcessSchemaInit,
TestHelperMultiStore, TestMultiStoreConcurrent_{InProcess,Subprocess,
CloseIsolation} all PASS. gofmt/go vet clean.

refs be-4oah9, be-yzowm

* deploy: release gate PASS for be-u8ldr (multiprocess dolt schema-init retry)

Gate evaluated against e2a7bbb90b (reviewed,
be-4oah9 round 3 PASS). All 7 criteria PASS: clean merge vs origin/main,
review PASS present, three-part acceptance criteria met, tests independently
re-run (7 PASS / 0 FAIL / 0 SKIP incl. adversarial ambient-port condition),
policy/lint lane clean, zero open HIGH findings, clean tree, single theme.

refs be-u8ldr, be-4oah9, be-yzowm, be-kafb

* deploy: correct gate criterion-7 evidence to cite full cumulative PR diff

The prior commit's criterion-7 line cited only the round-3 commit's
single-file stat. The actual PR diff (origin/main...e2a7bbb90b) spans
both files touched across the feature's full 4-commit lineage. Verdict
is unchanged (still single-theme PASS); this corrects the evidence
citation to match what actually ships.
2026-08-20 06:31:01 -07:00

6.9 KiB
Raw Permalink Blame History

Release gate — be-u8ldr (multiprocess dolt schema-init retry + test isolation fix)

Date: 2026-08-20 Deployer: beads/deployer Bead (deploy): be-u8ldr — needs-deploy: Fix: multiprocess tests INSERT without issues.description an (from:be-4oah9) Source bead (build): be-kafb — tdd_green fa825dcbd6 Review bead (round 1): be-yzowm — closed, request-changes → superseded by round 2/3 rework Review bead (round 2/3): be-4oah9 — closed, close_reason "pass", verdict: PASS, review_round: 3 Source commit: e2a7bbb90bcd1077bd785236a4915c899ed059d8 — "fix(dolt): remove dead decoy-listener test code, use retry-safe schema init under 10-way contention" (refs be-4oah9, be-yzowm, be-kafb) Provenance branch: builder/be-kafb — provenance only, not a push target (shared builder branch) Deploy branch (isolated, cut from source commit above): deploy/be-u8ldr-gate Base: origin/main @ d38ac728b581c8595fae36344ecca68830c7f3b5 ("fix(httpapi): make beads_dir and repo_root optional in ContextResponse (#5757)") Merge-base: ed382cbdb89cf7ba42b020e4927575dbf27e102e Merge-tree simulation: git merge-tree --write-tree origin/main e2a7bbb90b → tree 90320eb53df70b48ab8014c0ae231e673d0991ba, exit 0, zero conflict markers Ancestry scope check: assert_deploy_ancestry_scope origin/main e2a7bbb90b be-u8ldr be-4oah9 be-yzowm be-kafb → exit 0 (both commits between merge-base and source commit cite an accepted bead id; no .claude/** paths touched)

Pre-flight: already merged/PR'd?

DEPLOY_MODE=remote (origin and fork both resolve to github.com URLs). No PR URL recorded in be-u8ldr. Checked gh api repos/gastownhall/beads/commits/e2a7bbb90bcd1077bd785236a4915c899ed059d8/pulls → [] (no PR exists for this SHA). Checked git merge-base --is-ancestor e2a7bbb90b origin/main → not an ancestor (not already merged). Clean to proceed with the normal flow.

Verdict: PASS

Criteria walk

# Criterion Result Evidence
6 Branch diverges cleanly from main PASS git merge-tree --write-tree origin/main e2a7bbb90b succeeds, single merged tree, no conflicts. Source commit is 28 commits behind current origin/main, but none of those 28 touch internal/storage/dolt/initschema_multiprocess_test.go (git log --oneline SHA..origin/main -- <path> → empty). No self-rebase needed.
1 Review PASS present PASS be-4oah9 closed by beads/reviewer, close_reason "pass". Round-3 notes: style_findings all clean (gofmt/build/vet×2), security_findings — diff is a pure subset of round-1's already-walked 9-point OWASP review, zero new attack surface, no go.mod/go.sum change. verdict: PASS, review_round: 3.
2 Acceptance criteria met PASS Original three-part bug (title: "INSERT without issues.description and report OK on zero writes; dolt init has no identity"): (a) issues.description/design/acceptance_criteria/notes are TEXT NOT NULL with no default — INSERT now includes them (verified against internal/storage/schema/migrations/0001_create_issues.up.sql in be-yzowm round 1); (b) write-count mismatch now os.Exit(1)s instead of silently reporting OK (confirmed independently by round-2 reviewer); (c) dolt init --name/--email fix verified against baseline failure "empty ident name not allowed" (be-yzowm round 1). All three fixed and untouched by this round's diff — round 3 closes the sole remaining gap, the diff-owned test's own susceptibility to ambient dolt-server port/lock state.
3 Tests pass (documented CI-equivalent command, real counts) PASS Independently re-run by this deployer (not just trusting reviewer's report) at the exact source commit, disposable scratch worktree, DOCKER_HOST=unix:///run/user/1000/podman/podman.sock TESTCONTAINERS_RYUK_DISABLED=true go test -race -tags=integration,gms_pure_go -timeout=200s -count=1 -v -run '<pattern>' ./internal/storage/dolt/.... TestMultiProcessSchemaInit_DoltVerify: PASS×2 (clean env 8.95s; ambient BEADS_DOLT_SERVER_PORT=28231 set, replicating round-1/round-2's adversarial condition, 8.33s). Adjacent regression suite (TestMultiProcessSchemaInit, TestHelperMultiStore, TestMultiStoreConcurrent_{InProcess,Subprocess,CloseIsolation}): PASS×5. Totals: 7 PASS / 0 FAIL / 0 SKIP across both runs. gofmt -l, go build ./..., go vet -tags=integration,gms_pure_go ./internal/storage/dolt/...: all clean, exit 0.
3a Pre-existing-failure attribution N/A Not invoked — zero failures observed, nothing to attribute.
3b Policy/lint lane PASS make ci-pr-policy at source commit: exit 0. Doc-freshness, testing.Short() boundaries, workapi frontend boundary, .beads/issues.jsonl diff check, and openapi spec gate (incl. go test ./internal/httpapi/...) all reported PASS/succeeded; zero FAIL/ERROR lines in the full log.
4 No HIGH-severity findings open PASS Both round-2 BLOCKING findings (dead decoy-listener test code; migration-lock-timeout FAIL) were root-caused and fixed in round 3, independently re-verified structurally correct by the round-3 reviewer (not just trusted from builder's claim). waiver_ref: none needed. bd dep tree be-u8ldr / bd dep tree be-4oah9: no open child/dependent beads.
5 Feature branch / working tree clean PASS git status --short at the deployer worktree checked out to the source commit: empty (clean).
7 Single feature theme PASS Full PR diff git diff --stat origin/main...e2a7bbb90b: 2 files (internal/storage/dolt/initschema_multiprocess_test.go, internal/storage/dolt/multistore_multiprocess_test.go), 26 insertions(+), 5 deletions(-), across 4 commits (fa825dcbd tdd-green, dfdfa0440/d7233eb38 round-2 red/green rework, e2a7bbb90 round-3 fix). One coherent theme throughout: the bug title's own three symptoms (INSERT missing NOT NULL columns, zero-writes-reported-OK, dolt-init-missing-identity) plus the round-2/3 hardening of the diff-owned test's own reliability under ambient dolt-server state. No scope creep, no unrelated files.

Hand-off

  • DEPLOY_MODE=remote. Push target: headfork (https://github.com/quad341/beads-sec003-contrib.git) — not origin (gastownhall/beads; push disabled, fetch-only) and not builder/be-kafb (provenance only, a shared builder branch other beads may still commit to). fork (quad341/beads.git) is stale post-rename; headfork/prhead are the current remotes for this contributor rig.
  • PR: cross-repo quad341-sec003-contrib:deploy/be-u8ldr-gate → gastownhall:main. Run gc beads-contributor pre-pr-check before opening and clear every [BLOCK].
  • gastownhall/beads is upstream-only for this rig (contributor relationship, not maintainer). Per role instructions, job ends at opening the PR: do not merge, approve, or dismiss reviews there; never run maintainer-pr-review/mpr against it. Record the gate result + PR URL, close the bead, stand down.