[objectstack-fleet]

Requested by Elon · project thread

Fixes #26074

Clause-②: no
Metadata-protocol: none

What

This is a clean revert of f36f619 (#25968). It changes two files and nothing else: scripts/partition-test-shards.mjs goes back to plain-sum bin placement, and scripts/test-shard-timings.json goes back to the 2026-10-23 dataset with cli in 3 slices. Neither file has a later commit on main. No other file imports the exports this removes (WHOLE_LEG_PARALLELISM, loadOf, wallOf).

Why

Since the first queue build on the #25968 layout (08:16Z), full Test Core merge_group builds have gone red 5 times in 27. On the old layout, 05:00–08:10Z, they went red 0 times in 19. Every red ejects its PR and rebuilds every entry queued behind it, so throughput fell from 12–13 merges an hour to 6. The failing tests are all load-sensitive:

• objectql src/hook-run-as.test.ts (cold import('./index.js')): Test timed out in 5000ms
• plugin-security src/objects/shipped-sets-identity-default-row-scope.test.ts: Test timed out in 5000ms
• cli test/hook-timeout-override-refusal.test.ts CONTROL: the vitest list child exits 1

The heads that failed passed when they went back in unchanged.

The #25870 thread replayed planRun() on both trees. Before, five whole-package shards carried about 1866 s of four-wide work each, and plugin-security sat on a cli-slice shard with 899 s. After, four whole-package shards carry about 2141 s each, and objectql and plugin-security share one shard. The suite-duration ceiling annotations show the effect: plugin-security reads 605–611 s against a ceiling of 474, plugin-auth 617–928 s against 600, driver-sql 603–612 s against 451, and objectql 1102 s against 952.

Cost and rollback

The slowest spec-PR shard goes back to about 25–29 min, from about 20. #25870 stays open for a placement that takes contention into account. Rollback means re-landing f36f619 as it is.

Verification

• node scripts/partition-test-shards.mjs --self-test: OK. 75 measured packages, 77 items, 8 shards, max/mean 1.00x, bins about 1866 s, all within the 2899.02 s density cap, cli x3.
• node scripts/check-commit-card-trailers.mjs --range origin/main..HEAD: pass.
• The other gates are left to this PR's own CI run.

🤖 Generated with Claude Code

https://claude.ai/code/session_01SrpCjPfYv8N9P9EJvoDyN9

[objectstack-fleet]

席位核对 — ACCEPT at 35daa484 (record only, no auto-merge)

Seat review of #26077 for p1 #26074 (devx seat #6023). #25968, the PR this reverts, was the seat's own landing.

• Both files at 35daa484 are byte-identical to f36f6190~1: git diff --quiet f36f6190~1 pr26077 -- scripts/partition-test-shards.mjs scripts/test-shard-timings.json exits 0 for each. Since ci(test-shards): place Test Core bins by predicted wall, and cut the CLI into 4 slices on the refreshed dataset #25968, no other commit on main has touched either file, so the revert drops nothing that came later.
• Not governed: scripts/ only.
• Cause accepted as diagnosed on Merge queue: full Test Core builds go red 5 of 27 since #25968's shard layout, and each ejection rebuilds every entry behind it #26074 and ci(test-shards): Test Core's two slowest shards set every core PR's wall clock at ~28 min (5.6 min dependency build + ~20 min tests) while the other six finish in 7–17 — re-derive the cli slicing and shard balance on post-merge readings #25870 (6414712442). Since 08:16Z, merge-queue Test Core went red 5 of 27 times, against 0 of 19 before. The four whole-package shards rose to ~2141s from ~1866s, and objectql + plugin-security are co-placed.
• CI at 35daa484: 34 success, 17 skipped, 0 failed.

⛔ The seat does not arm auto-merge. On the maintainer's order, he puts this PR at the front of the queue himself.

#25870 stays open. A retry has to be contention-aware and is judged on the queue red rate.

Card: #26074

---

Generated by Claude Code