Files
PaperClipAI/scripts/__tests__/run-vitest-stable-shard.test.mjs
00a24d7e8f ci: split general-server tests into five shards with refreshed durations (#10925)
## Thinking Path

> - Paperclip is the open source app people use to manage AI agents for
work.
> - The PR workflow runs the server vitest suite across sharded runners
because the suite is pinned to one worker.
> - In successful PR run 30930345729 (2026-08-04), shard `server (3/4)`
took 311 seconds of wall time and was the slowest check in the run.
> - The suite has grown to about 946 seconds of serial vitest time, but
the duration manifest was last sampled on 2026-08-01 at about 882
seconds.
> - This pull request refreshes the per-suite duration manifest from
that run's logs and splits the lane into five shards.
> - The benefit is a shorter PR critical path: each shard carries about
196 seconds of suite time, level with the other lanes.

## Linked Issues or Issue Description

Refs #10663 (previous split of this lane into four shards).
Related: #10923 splits the separate serialized-suites lane into five
shards. Both PRs touch `.github/workflows/pr.yml` in different matrix
blocks; whichever merges second needs a trivial rebase.

**What existing behavior does this improve?**

The `general-server` vitest lane runs in four shards with a duration
manifest sampled on 2026-08-01.

**Current behavior**

In PR run 30930345729, shard 3/4 ran for 311 seconds (273 seconds in the
test step) and was the longest check in the run. The suite now totals
about 946 seconds of serial vitest time.

**Proposed behavior**

Run the same suite set in five shards, balanced with a per-suite
duration manifest refreshed from that run's shard logs (279 suites
measured by diffing consecutive completion timestamps).

**Reason and benefit**

The refreshed LPT partition balances at about 196 seconds of suite time
per shard (about 240 seconds per job), level with the other PR lanes. No
test coverage is lost.

**Breaking changes**

None. The change only alters the CI partition size and the duration
manifest.

## What Changed

- Bump the `general-server` shard matrix in `.github/workflows/pr.yml`
from four to five shards.
- Refresh `scripts/general-server-shard-durations.json` from the
2026-08-04 run's shard logs.
- Update `SHARD_COUNT` in
`scripts/__tests__/run-vitest-stable-shard.test.mjs` to five.

## Verification

- `node --test scripts/__tests__/run-vitest-stable-shard.test.mjs` — 9/9
pass, including the complete non-overlapping partition proof and the
duration-balance check.
- `node --test scripts/__tests__/release-verify-workflow.test.mjs` — 2/2
pass.
- `node --test scripts/__tests__/e2e-shard.test.mjs` — 7/7 pass.
- A 5-way dry-run partition covers all suites exactly once with equal
projected weights.

## Risks

- Low risk. The change only alters CI partition size and duration
weights; the suite set is unchanged.
- One more runner is used per PR run for this lane.
- Stale duration weights degrade gracefully: suites missing from the
manifest get the median weight.

## Model Used

- Claude (Anthropic), Claude Code CLI, model ID `claude-fable-5`,
extended thinking with tool use enabled.

## Checklist

- [x] I have included a thinking path that traces from project context
to this change
- [x] I have specified the model used (with version and capability
details)
- [x] I have checked ROADMAP.md and confirmed this PR does not duplicate
planned core work
- [x] I have searched GitHub for duplicate or related PRs and linked
them above
- [x] I have either (a) linked existing issues with `Fixes: #` / `Closes
#` / `Refs #` OR (b) described the issue in-PR following the relevant
issue template
- [x] I have not referenced internal/instance-local Paperclip issues or
links (only public GitHub `#NNN` / `github.com/paperclipai/paperclip`
URLs)
- [x] My branch name describes the change (e.g. `docs/...`, `fix/...`)
and contains no internal Paperclip ticket id or instance-derived details
- [x] I have run tests locally and they pass
- [x] I have added or updated tests where applicable
- [x] I have updated relevant documentation to reflect my changes
(workflow comments explain the new shard math)
- [x] I have considered and documented any risks above
- [ ] All Paperclip CI gates are green
- [ ] Greptile is 5/5 with no open P2s, recommendations, or follow-ups
- [x] I will address all Greptile and reviewer comments before
requesting merge

Co-authored-by: Claude <claude@paperclip.ing>
Co-authored-by: Paperclip <noreply@paperclip.ing>
2026-08-05 16:44:29 -07:00

161 lines
7.0 KiB
JavaScript

import assert from "node:assert/strict";
import { spawnSync } from "node:child_process";
import path from "node:path";
import { fileURLToPath } from "node:url";
import test from "node:test";
import {
defaultSuiteWeight,
loadShardDurations,
partitionGeneralServerSuites,
} from "../general-server-shard.mjs";
const repoRoot = path.resolve(path.dirname(fileURLToPath(import.meta.url)), "..", "..");
const script = path.join(repoRoot, "scripts", "run-vitest-stable.mjs");
const durationsManifest = path.join(repoRoot, "scripts", "general-server-shard-durations.json");
function dryRun(args) {
const result = spawnSync(process.execPath, [script, ...args, "--dry-run"], {
cwd: repoRoot,
encoding: "utf8",
});
return result;
}
function dryRunJson(args) {
const result = dryRun(args);
assert.equal(result.status, 0, `expected success for ${args.join(" ")}: ${result.stderr}`);
return JSON.parse(result.stdout);
}
const SHARD_COUNT = 5;
const SERIALIZED_SHARD_COUNT = 5;
test("the serialized shards form a complete, non-overlapping partition", () => {
const shards = Array.from({ length: SERIALIZED_SHARD_COUNT }, (_, index) =>
dryRunJson(["--mode", "serialized", "--shard-index", String(index), "--shard-count", String(SERIALIZED_SHARD_COUNT)]),
);
const total = shards[0].serializedSuiteCount;
const selected = shards.flatMap((shard) => shard.selectedSerializedSuites);
assert.equal(selected.length, total, "every serialized suite must be selected exactly once");
assert.equal(new Set(selected).size, total, "serialized shards must not overlap");
});
test("the general-server shards form a complete, non-overlapping partition", () => {
const shards = Array.from({ length: SHARD_COUNT }, (_, index) =>
dryRunJson(["--mode", "general", "--group", "general-server", "--shard-index", String(index), "--shard-count", String(SHARD_COUNT)]),
);
const total = shards[0].generalServerSuiteCount;
assert.ok(total > 0, "expected a non-empty general-server suite set");
const seen = new Set();
let selectedTotal = 0;
for (const shard of shards) {
assert.equal(shard.generalServerSuiteCount, total, "suite count must be stable across shards");
for (const file of shard.selectedGeneralServerSuites) {
assert.ok(!seen.has(file), `suite assigned to more than one shard: ${file}`);
seen.add(file);
selectedTotal += 1;
}
}
// Every suite runs exactly once: union covers the whole set with no overlap.
assert.equal(selectedTotal, total, "every suite must be selected exactly once");
assert.equal(seen.size, total, "union of shards must cover the whole suite set");
});
test("a route/authz suite never leaks into the general-server shards", () => {
const shard = dryRunJson(["--mode", "general", "--group", "general-server", "--shard-index", "0", "--shard-count", SHARD_COUNT.toString()]);
for (const file of shard.selectedGeneralServerSuites) {
assert.ok(
!/[^/]*(?:route|routes|authz)[^/]*\.test\.ts$/.test(file),
`route/authz suite must stay in the serialized lane, not general-server: ${file}`,
);
}
});
test("shard flags are rejected for the parallel workspace groups", () => {
const result = dryRun(["--mode", "general", "--group", "general-workspaces-a", "--shard-index", "0", "--shard-count", "3"]);
assert.notEqual(result.status, 0, "workspace groups must not accept shard flags");
});
test("duration-aware partition balances skewed weights better than round-robin", () => {
// Round-robin puts all three heavy suites on shard 0 (indexes 0, 3, 6).
const files = ["a", "b", "c", "d", "e", "f", "g", "h", "i"];
const durations = { a: 30000, d: 30000, g: 30000, b: 100, c: 100, e: 100, f: 100, h: 100, i: 100 };
const shards = partitionGeneralServerSuites(files, 3, durations);
const totals = shards.map((shard) => shard.totalWeight);
const maxTotal = Math.max(...totals);
const minTotal = Math.min(...totals);
assert.ok(
maxTotal - minTotal <= 200,
`expected near-even shard weights, got ${totals.join(", ")}`,
);
assert.equal(
shards.flatMap((shard) => shard.files).sort().join(","),
files.join(","),
"partition must cover every file exactly once",
);
});
test("the partition is deterministic for identical inputs", () => {
const files = Array.from({ length: 50 }, (_, index) => `suite-${index}.test.ts`);
const durations = Object.fromEntries(files.map((file, index) => [file, (index * 37) % 5000]));
const first = partitionGeneralServerSuites(files, 3, durations);
const second = partitionGeneralServerSuites(files, 3, durations);
assert.deepEqual(first, second, "same inputs must always produce the same partition");
});
test("suites missing from the manifest get the median weight", () => {
assert.equal(defaultSuiteWeight({ a: 100, b: 300, c: 900 }), 300);
assert.equal(defaultSuiteWeight({ a: 100, b: 300, c: 500, d: 900 }), 400);
assert.equal(defaultSuiteWeight({}), 1000, "empty manifest falls back to a fixed weight");
});
test("a missing or malformed manifest degrades to uniform weights", () => {
assert.deepEqual(loadShardDurations(path.join(repoRoot, "scripts", "no-such-manifest.json")), {});
const files = ["a", "b", "c", "d"];
const shards = partitionGeneralServerSuites(files, 2, {});
assert.equal(shards[0].files.length + shards[1].files.length, files.length);
assert.equal(Math.abs(shards[0].files.length - shards[1].files.length), 0);
});
test("the checked-in manifest loads and covers most of the current suite set", () => {
const durations = loadShardDurations(durationsManifest);
assert.ok(Object.keys(durations).length > 0, "manifest must parse to a non-empty duration map");
const shard = dryRunJson(["--mode", "general", "--group", "general-server", "--shard-index", "0", "--shard-count", "1"]);
const currentFiles = shard.selectedGeneralServerSuites;
const known = currentFiles.filter((file) => durations[file] !== undefined).length;
assert.ok(
known / currentFiles.length >= 0.5,
`manifest is stale: only ${known} of ${currentFiles.length} suites have recorded durations — regenerate it from a recent PR run (see the manifest's $comment)`,
);
});
test("the real shard partition is duration-balanced", () => {
const durations = loadShardDurations(durationsManifest);
const fallback = defaultSuiteWeight(durations);
const shards = Array.from({ length: SHARD_COUNT }, (_, index) =>
dryRunJson(["--mode", "general", "--group", "general-server", "--shard-index", String(index), "--shard-count", String(SHARD_COUNT)]),
);
const totals = shards.map((shard) =>
shard.selectedGeneralServerSuites.reduce((sum, file) => sum + (durations[file] ?? fallback), 0),
);
const maxTotal = Math.max(...totals);
const minTotal = Math.min(...totals);
// LPT keeps the spread within the heaviest single suite; use that as the bound.
const heaviest = Math.max(...Object.values(durations));
assert.ok(
maxTotal - minTotal <= heaviest,
`shard weight spread ${maxTotal - minTotal}ms exceeds heaviest suite ${heaviest}ms: ${totals.join(", ")}`,
);
});