Skip to content

[e2e] Re-enable concurrent execution of the e2e suite - #3658

Merged
VaguelySerious merged 5 commits into
mainfrom
alangenfeld/e2e-concurrent-v2
Aug 20, 2026
Merged

[e2e] Re-enable concurrent execution of the e2e suite#3658
VaguelySerious merged 5 commits into
mainfrom
alangenfeld/e2e-concurrent-v2

Conversation

@alangenfeld

@alangenfeldalangenfeld commented Aug 19, 2026•
edited
Loading

Copy link
Copy Markdown
Collaborator

Summary & Motivation

Serial execution was the dominant wall-clock cost per matrix entry (~22 of 24 minutes on the Vercel lanes), and the blockers that forced it in 78048e0 are fixed.

  • Per-test harness state (name, tracked runs, skip) binds through AsyncLocalStorage instead of module globals, so a failing test dumps its own diagnostics and a mid-body conformance gate skips the right test. The suite's test is a createTaskCollector wrapper that delegates to the enclosing suite's collector, which is both how the binding gets a direct call stack to the handler and how describe.concurrent inheritance reaches each test.
  • Each lane ends with a load summary — test and CLI child counts, peak concurrency of each, median test wall time, CLI share of summed test time, ten slowest tests — since under concurrency the question is where per-test latency went, not pass/fail.
  • maxConcurrency is tunable via WORKFLOW_E2E_MAX_CONCURRENCY, default unchanged at 5. Higher values starve CLI children on lanes that also host the app and Postgres.
  • RACE_WINNER_MAX_DURATION_MS goes 5s → 8s; queue latency under concurrency pushed observed winner durations to ~6.5s, still far below the 10s loser.

A test that genuinely cannot share a deployment can opt out with test.sequential.

Builds on VaguelySerious's investigation in #2083.

Test Plan

Added a unit test for interleaved per-test attribution; five consecutive full-matrix CI runs at maxConcurrency=5 were green (162/162) with Vercel-prod lanes at 5-7 min against 22-25 min serial and median test time unchanged, and a maxConcurrency=8 probe established 5 as the value safe on every lane type.

@changeset-bot

changeset-botBot commented Aug 19, 2026•
edited
Loading

Copy link
Copy Markdown

⚠️ No Changeset found

Latest commit: c35ea2c

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

@vercel

vercelBot commented Aug 19, 2026•
edited
Loading

Copy link
Copy Markdown
Contributor

The latest updates on your projects. Learn more about Vercel for GitHub.

ProjectDeploymentActionsUpdated (UTC)
example-nextjs-workflow-turbopackReadyReadyPreviewAug 19, 2026 5:45pm
example-nextjs-workflow-webpackReadyReadyPreviewAug 19, 2026 5:45pm
example-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-astro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-express-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-fastify-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-hono-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nestjs-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nitro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nuxt-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-python-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-sveltekit-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-tanstack-start-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-vite-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workflow-docsReadyReadyPreview, v0Aug 19, 2026 5:45pm
workflow-swc-playgroundReadyReadyPreviewAug 19, 2026 5:45pm
workflow-tarballsReadyReadyPreviewAug 19, 2026 5:45pm
workflow-webReadyReadyPreviewAug 19, 2026 5:45pm

@alangenfeld
[e2e] Re-enable concurrent execution of the e2e suite
fc94f67
Serial execution has been the dominant wall-clock cost per matrix entry
since concurrency was disabled before conf (78048e0): ~128 tests at
~22 of 24 minutes on the Vercel lanes, and lately the slowest lane
cannot finish under its 30-minute job timeout on a slow runner day at
all. #2083 measured the concurrent suite at ~3x job wall-clock (4-5x on
the vitest phase) and identified what broke; its blockers are now fixed:
world-local writeExclusive is atomic (write-then-link), abort-fetch
tests are hermetic (#3618), the fibonacci tree fits the scheduler
(#3619), and source-map assertions are positive-only (#3620).
What this change adds is concurrency-safe per-test attribution. The
harness tracked runs and test names in module globals reset by a
beforeEach - under concurrency every test clobbered every other's
state, so a failing test dumped an unrelated sibling's diagnostics.
vitest's getCurrentTest() cannot substitute: it is a plain module
variable, wrong after any await. Instead an auto fixture - the one
place that receives the test's own context unambiguously - binds a
per-test state (name, tracked runs, the test's own skip) via
AsyncLocalStorage around each test body, and trackRun /
recordInfraEvent / requireFixture read it ambiently with no call-site
changes. The conformance gates skip through the bound state's skip, so
a mid-body requireFixture skips the right test. Sequential suites
(dev, agent, region) keep setupRunTracking's module-global fallback.
Full suite passes 137/137 concurrently against a local dev server in
under 2 minutes. A test that genuinely cannot share a deployment can
opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
[e2e] Bind per-test state via a task collector, not a fixture
f437a21
CI on the first concurrent run caught the fixture design failing in the
one place skip-correctness is load-bearing: vitest resolves fixtures in
a separate async context, so an AsyncLocalStorage store bound around
use() never reaches the test body. Every ambient read silently fell to
the module fallback - conformance-gated tests on the Python lane ran
instead of skipping and hard-failed on fixtures missing from the
manifest ('declared in e2e-conformance.json but not in the deployed
manifest').
The suite's test is now built with createTaskCollector, mirroring
vitest's own argument normalization, and the wrapper binds the state
around the handler call itself - a direct call stack, so propagation is
guaranteed rather than assumed. Validated by a collector-pattern ALS
experiment (per-test store after awaits, mid-body skip targets self,
each/skip/todo chains), a conformance-file simulation reproducing the
Python failure shape (fixture-missing tests now skip with the right
note), and a 137/137 concurrent local run.
Also raises RACE_WINNER_MAX_DURATION_MS 5s -> 8s: the winner takes 1s,
the loser 10s, and the bound only has to sit clearly below the loser -
concurrent-suite queue latency pushed observed winner durations to
~6.5s on loaded local-dev lanes (vite, first concurrent CI run), so 5s
flaked without catching anything 8s misses.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
alangenfeldforce-pushed the alangenfeld/e2e-concurrent-v2 branch from 0556953 to f437a21CompareAugust 19, 2026 15:05
@github-actions

github-actionsBot commented Aug 19, 2026•
edited
Loading

Copy link
Copy Markdown
Contributor

📊 Workflow Benchmarks

commit c35ea2c · Wed, 19 Aug 2026 18:18:12 GMT · run logs

Backend: vercel · app: nextjs-turbopack

MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1380 (+381%) 🔻1517 🔴 (+30%) 🔻1557 🔴 (+16%) 🔻1624 🔴 (+3.9%)30
TTFSstream311 (+65%) 🔻1510 🔴 (+35%) 🔻1575 🔴 (+38%) 🔻1913 🔴 (+64%) 🔻30
TTFShook + stream1625 (+25%) 🔻1970 🔴 (+43%) 🔻2048 🔴 (+45%) 🔻2524 🔴 (+47%) 🔻30
Fan-out TTFSPromise.all(100 steps)1152 (+4.3%)2804 (+22%) 🔻3163 (+14%)3207 (+16%) 🔻10
Fan-out TTLSPromise.all(100 steps)5366 (+1.6%)8639 (+30%) 🔻10891 (+58%) 🔻19609 (+149%) 🔻10
STSO1020 steps (inline)117 (-11%)164 (-38%) 💚190 (-38%) 💚365 (-28%) 💚1019
WO1020 steps166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚1
CRTTfirst chunk (pooled)112 (+33%) 🔻186 (+46%) 🔻300 (-4.8%)436 (-30%) 💚28

Streams

ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)172 (+52%)176 (+7%)254 (+9%)457 (-2%)159 (-1%)10
size sweep (100/s, 160B-12KB)149 (+39%)165 (+7%)324 (+15%)518 (-24%)151 (-6%)10
replay gateway-gpt-5.4-nano-2000t (1x)126 (+13%)158 (-43%)240 (-67%)719 (-52%)353 (-41%)3
replay eve-gpt-5.6-sol-2000t (1x)135 (-14%)146 (+4%)186 (-3%)743 (+18%)523 (+7%)2
replay eve-gpt-5.6-sol-2000t (2x)138 (-73%)247 (-39%)355 (-87%)729 (-83%)569 (-6%)3
📈 STSO distribution vs main (inline / queue-hop histograms)

1020 steps (inline)

Cumulative STSO time: main 247209ms → this run 164994ms (Δ -82215ms, -33%)

100-150 ms █░░░░░░░░░░░░░░░░░░░░░░┃ main 6 this 474 +468
150-200 ms ██████████░░░░░░░░░░░░┃ main 205 this 464 +259
200-250 ms █┃██████████████████████ main 481 this 45 -436
250-300 ms ┃██████████ main 217 this 15 -202
300-350 ms ┃██ main 58 this 7 -51
350-400 ms ┃ main 20 this 8 -12
400-450 ms ┃ main 12 this 3 -9
450-500 ms ┃ main 9 this 1 -8
500-550 ms ┃ main 4 this 1 -3
550-600 ms ┃ main 2 this 0 -2
600-650 ms ┃ main 4 this 1 -3
850-900 ms ┃ main 1 this 0 -1
📈 CRTT drill-down vs main (RTT distributions & profiles)
variant RTT 1ms→5s+ avg p50 p90 p99 n
control ······▂█▂···· 148.6 (+19%) 137 (+27%) 254 (+9%) 457 (-2%) 3000
sweep ······▃█▂▁··· 145.5 (+10%) 129 (+15%) 324 (+15%) 518 (-24%) 3000
gw 1x ·····▁▄█▁▁··· 132.2 (-25%) 116 (+5%) 240 (-67%) 719 (-52%) 5295
eve 1x ·····▁▅█▁▁··· 133.2 (+7%) 113 (+5%) 186 (-3%) 743 (+18%) 5186
eve 2x ·····▁▂█▃▁··· 173.2 (-55%) 139 (-19%) 355 (-87%) 729 (-83%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):

control █▄▃▂▄▂▄▁▁▂ 131–192ms
sweep ▇▃▂▁▃▃▄█▂▂ 124–181ms
gw 1x ▃█▄▂▂▄▁▂▃▄ 110–178ms
eve 1x ▂▃█▃▁▂▅▄▃█ 101–182ms
eve 2x █▁▂▂▁▄▅▅▄▅ 127–257ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):

sweep █▆▃▁▄▅█ 142–148ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):

control ▁▇▆▅▇▄█▄▆▄ 27–47ms
sweep ▁▆▆▆█▇▇▇▅▄ 31–61ms
gw 1x ▄▆▄▁▄█▃▄▃▅ 32–45ms
eve 1x ▄▆▆▇▁▄█▆▃█ 20–29ms
eve 2x █▁▃▂▃▅▃▁▄▅ 18–35ms
📜 Previous results (2)

c35ea2c

Wed, 19 Aug 2026 17:11:18 GMT · run logs

vercel / nextjs-turbopack

MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1255 (+337%) 🔻1314 🔴 (+13%)1358 🔴 (+1.4%)1383 🔴 (-12%)30
TTFSstream241 (+28%) 🔻1334 🔴 (+20%) 🔻1399 🔴 (+23%) 🔻1681 🔴 (+44%) 🔻30
TTFShook + stream367 (-72%) 💚1748 🔴 (+27%) 🔻1778 🔴 (+26%) 🔻1952 🔴 (+14%)30
Fan-out TTFSPromise.all(100 steps)1392 (+26%) 🔻2936 (+28%) 🔻3132 (+13%)3151 (+14%)10
Fan-out TTLSPromise.all(100 steps)5963 (+13%)7084 (+6.4%)7237 (+5.1%)8436 (+7.0%)10
STSO1020 steps (inline)111 (-16%) 💚272 (+3.4%)320 (+5.3%)495 (-2.0%)1019
WO1020 steps245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)1
CRTTfirst chunk (pooled)98 (+17%) 🔻169 (+33%) 🔻398 (+26%) 🔻824 (+31%) 🔻28

Streams

ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)135 (+19%)434 (+163%)618 (+166%)1102 (+137%)617 (+287%)10
size sweep (100/s, 160B-12KB)149 (+39%)344 (+123%)540 (+92%)1044 (+54%)288 (+79%)10
replay gateway-gpt-5.4-nano-2000t (1x)127 (+13%)169 (-39%)290 (-60%)750 (-50%)524 (-13%)3
replay eve-gpt-5.6-sol-2000t (1x)206 (+32%)151 (+8%)246 (+28%)501 (-20%)380 (-22%)2
replay eve-gpt-5.6-sol-2000t (2x)153 (-70%)191 (-53%)315 (-89%)676 (-84%)444 (-27%)3

f437a21

Wed, 19 Aug 2026 15:26:31 GMT · run logs

vercel / nextjs-turbopack

MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1341 (+367%) 🔻1586 🔴 (+36%) 🔻1690 🔴 (+26%) 🔻1885 🔴 (+21%) 🔻30
TTFSstream362 (+92%) 🔻1460 🔴 (+31%) 🔻1538 🔴 (+35%) 🔻1747 🔴 (+49%) 🔻30
TTFShook + stream1593 (+22%) 🔻1872 🔴 (+36%) 🔻1946 🔴 (+38%) 🔻2005 🔴 (+17%) 🔻30
Fan-out TTFSPromise.all(100 steps)1248 (+13%)2227 (-3.2%)2780 (+0.5%)3913 (+41%) 🔻10
Fan-out TTLSPromise.all(100 steps)5426 (+2.7%)11957 (+80%) 🔻22902 (+233%) 🔻24329 (+209%) 🔻10
STSO1020 steps (inline)111 (-16%) 💚164 (-38%) 💚187 (-38%) 💚285 (-44%) 💚1019
WO1020 steps159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚1
CRTTfirst chunk (pooled)96 (+14%)204 (+61%) 🔻341 (+8.3%)765 (+22%) 🔻28

Streams

ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)141 (+25%)184 (+12%)398 (+72%)735 (+58%)152 (-5%)10
size sweep (100/s, 160B-12KB)168 (+57%)155 (+1%)206 (-27%)408 (-40%)126 (-22%)10
replay gateway-gpt-5.4-nano-2000t (1x)316 (+182%)167 (-40%)225 (-69%)429 (-71%)293 (-51%)3
replay eve-gpt-5.6-sol-2000t (1x)161 (+3%)166 (+19%)217 (+13%)556 (-11%)270 (-45%)2
replay eve-gpt-5.6-sol-2000t (2x)162 (-68%)201 (-50%)264 (-91%)537 (-87%)177 (-71%)3
ℹ️ Metric definitions & methodology

Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.

The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.

The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.

Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.

Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)

Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles

Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041

🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600

All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).

Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.

@github-actions

github-actionsBot commented Aug 19, 2026•
edited
Loading

Copy link
Copy Markdown
Contributor

🧪 E2E Test Results

✅ All tests passed

⚠️ Flaky E2E Tests (passed on retry)

These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.

  • cancelRun via CLI - cancelling a running workflow (astro)
  • cancelRun via CLI - cancelling a running workflow (vite)
  • deploymentId: 'latest' is a no-op in non-Vercel worlds (nitro)
  • hookGetConflictThenStepParallelWorkflow - hook.getConflict() continuation step runs alongside other steps (nextjs-webpack)

🛠 Infra Events (absorbed by the harness)

Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.

  • cold-start-warmup · suite warmup (tanstack-start) · at 18:14:32Z · abandoned wrun_01M0DKNYTD6N9KWDRTEMB63XWM
  • run-pickup-stall · hookGetConflictWorkflow - hook.getConflict() resolves with the conflicting run when token is already registered (nuxt) · at 18:15:38Z · abandoned wrun_01M0DKRB1H8KCG69A7VFN9V5XN

E2E Test Summary

Summary
PassedFailedSkippedTotal
✅ ▲ Vercel Production347407384212
✅ 💻 Local Development381005584368
✅ 📦 Local Production381005584368
✅ 🐘 Local Postgres381005584368
✅ 🪟 Windows31200312
✅ 🌐 Cross-language Conformance90128137
✅ vercel-multi-region270027
Total152520254017792
Details by Category

✅ ▲ Vercel Production

AppPassedFailedSkipped
✅ astro-node128028
✅ astro-quickjs128028
✅ example-node128028
✅ example-quickjs128028
✅ express-node128028
✅ express-quickjs128028
✅ fastify-node128028
✅ fastify-quickjs128028
✅ hono-node128028
✅ hono-quickjs128028
✅ nest-node128028
✅ nest-quickjs128028
✅ nextjs-turbopack-node15303
✅ nextjs-turbopack-quickjs15303
✅ nextjs-webpack-node15303
✅ nextjs-webpack-quickjs15303
✅ nitro-node128028
✅ nitro-quickjs128028
✅ nuxt-node128028
✅ nuxt-quickjs128028
✅ python-node80148
✅ sveltekit-node14709
✅ sveltekit-quickjs14709
✅ tanstack-start-node128028
✅ tanstack-start-quickjs128028
✅ vite-node128028
✅ vite-quickjs128028

✅ 💻 Local Development

AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026

✅ 📦 Local Production

AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026

✅ 🐘 Local Postgres

AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026

✅ 🪟 Windows

AppPassedFailedSkipped
✅ nextjs-turbopack-node15600
✅ nextjs-turbopack-quickjs15600

✅ 🌐 Cross-language Conformance

AppPassedFailedSkipped
✅ python90128

✅ vercel-multi-region

AppPassedFailedSkipped
✅ nextjs-turbopack2700

📋 View full workflow run

@github-actions

github-actionsBot commented Aug 19, 2026•
edited
Loading

Copy link
Copy Markdown
Contributor

Sim World

Simulated world deterministic testing for races. Traces

🟠 world-sim scenario book — 1 fail of 41 total

fence=per-spec

scenariooutcomeeventsvirtreplayviolations
✅ smoke-no-stepscompleted30msok0
✅ smoke-one-stepcompleted60msok0
✅ hook-at-step-startedcompleted120msok0
✅ hook-at-step-completedcompleted120msok0
✅ hook-at-hook-createdcompleted120msok0
✅ deadline-hook-winscompleted71.0hok0
✅ deadline-expirescompleted71.0hok0
✅ long-sleepcompleted1130.0dok0
✅ hook-never-arrivesstalled30msskipped0
✅ step-retries-twicecompleted102.0sok0
✅ parallel-stepscompleted90msok0
✅ hook-on-execution-statecompleted120msok0
✅ peek-hook-before-branchcompleted120msok0
✅ peek-hook-after-branchcompleted120msok0
✅ peek-hook-at-registrationcompleted120msok0
✅ race-hook-before-probecompleted120msok0
✅ race-hook-after-probecompleted120msok0
✅ race-duplicate-deliverycompleted130msok0
✅ attr-hook-before-stepcompleted110msok0
✅ attr-hook-after-stepcompleted110msok0
✅ attr-from-step-bodycompleted130msok0
✅ fork-hook-after-timeoutcompleted141.0mok0
✅ fork-hook-before-timeoutcompleted141.0mok0
✅ count-hook-after-timeoutcompleted171.0mok0
✅ count-hook-before-timeoutcompleted201.0mok0
✅ stale-read-step-count-forkcompleted201.0mok0
✅ stale-read-equal-step-countscompleted141.0mok0
✅ step-vs-step-forkcompleted120msok0
✅ step-vs-step-fork-fencedcompleted120msok0
✅ fence-catches-benign-directioncompleted125msok0
✅ in-flight-before-decisioncompleted171.0mok0
❌ in-flight-before-decision-countedcompleted171.0mok0
✅ in-flight-after-decisioncompleted192.0mok0
✅ stale-read-step-count-fork-fencedcompleted201.0mok0
✅ fork-hook-winscompleted131.0mok0
✅ fork-timeout-winscompleted131.0mok0
✅ unclaimed-payload-under-forkcompleted171.0mok0
✅ claimed-payload-under-forkcompleted171.0mok0
✅ writers-independent-step-bodiescompleted120msok0
✅ writers-scripted-tempocompleted120msok0
✅ cancel-mid-stepcancelled70msskipped0

Full trace: world-sim.txt

@VaguelySeriousVaguelySerious mentioned this pull request Aug 19, 2026
Merged
@alangenfeld
[e2e] Delegate to the suite collector so describe.concurrent applies
ef36a7e
The previous commit's collector called getCurrentSuite().task() directly,
which skips the suite collector's Object.assign({}, suiteOptions,
options) - the merge that carries describe.concurrent onto each test. The
suite therefore still ran sequentially, and CI proved it: every Vercel
Prod lane matched the serial baseline minute-for-minute
(nextjs-turbopack quickjs 25m vs 25m), so a fully green run said nothing
about concurrency. The same slowdown showed up locally as a full-suite
run taking 6m15s where the earlier concurrent one took 1m58s, which I
had misread as server slowness.
The collector now wraps the handler and hands the call to
getCurrentSuite().test.fn - exactly what vitest's own top-level test
does - so suite-option inheritance is vitest's code path again, while
the wrapper still owns the handler call stack that makes the
AsyncLocalStorage binding reach the test body.
Verified with an experiment asserting both properties at once: four
300ms tests plus a 600ms observer finish in 602ms (1800ms if
sequential), max-in-flight > 1, and each test still reads its own store
after awaits.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
vercel[bot]
vercelBot reviewed Aug 19, 2026
View reviewed changes
Comment threadpackages/core/e2e/e2e.test.ts
@alangenfeld
[e2e] Actually call the load-observability hooks
c35ea2c
The previous commit added the counters and the summary but left the call
sites out of the handler wrapper and afterAll, so nothing was collected
and no summary printed. Caught by review.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@github-actions

github-actionsBot commented Aug 20, 2026

Copy link
Copy Markdown
Contributor

No backport to stable for 8c29b85 (AI decision).

This is test-harness enhancement work, not a fix: it re-enables describe.concurrent for the e2e suite to cut CI wall clock, adds a new load-summary reporting facility, a new WORKFLOW_E2E_MAX_CONCURRENCY config knob, and reworks per-test state onto AsyncLocalStorage. Nothing here repairs broken or flaky behavior on stable — it trades a known-serial, known-green suite for a concurrency model whose safe value was tuned against main's lanes, which is exactly the kind of risk a maintenance branch is avoiding. The RACE_WINNER_MAX_DURATION_MS 5s→8s bump is the only flake-adjacent piece, and it is only needed because of the concurrency this commit introduces.

To override, re-run the Backport to stable workflow manually via workflow_dispatch and paste this commit SHA into the ref input:

8c29b85322dc2279b6274391cd14a1483eb969c3

Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@vercelvercel[bot]vercel[bot] left review comments
@TooTallNateTooTallNateTooTallNate approved these changes

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@alangenfeld@TooTallNate@VaguelySerious
, 'i'); if (__m === '*' || __re.test(location.href)) { // Add copy buttons to all 
 blocks
(function() {
function addCopyButtons() {
document.querySelectorAll('pre code').forEach(function(codeBlock) {
if (codeBlock.parentElement.hasAttribute('data-copy-added')) return;
codeBlock.parentElement.setAttribute('data-copy-added', 'true');
var btn = document.createElement('button');
btn.textContent = 'Copy';
btn.style.cssText = 'position:absolute;top:4px;right:4px;padding:2px 8px;font-size:11px;background:#4ecdc4;border:none;border-radius:4px;color:#1a1a2e;cursor:pointer;opacity:0.7;transition:opacity 0.2s;';
btn.onmouseover = function() { this.style.opacity = '1'; };
btn.onmouseout = function() { this.style.opacity = '0.7'; };
btn.onclick = function() {
navigator.clipboard.writeText(codeBlock.textContent).then(function() {
btn.textContent = 'Copied!';
setTimeout(function() { btn.textContent = 'Copy'; }, 1500);
});
};
codeBlock.parentElement.style.position = 'relative';
codeBlock.parentElement.appendChild(btn);
});
}
addCopyButtons();
// Re-run on dynamic content
var observer = new MutationObserver(addCopyButtons);
observer.observe(document.body, { childList: true, subtree: true });
})();
}
} catch(__e) { console.warn('[Userscript:Add Copy Buttons to Code Blocks]', __e); }
})();
(function(){
try {
var __m = "github.com";
var __re = new RegExp('^' + "github\\.com" + '
[e2e] Re-enable concurrent execution of the e2e suite by alangenfeld · Pull Request #3658 · vercel/workflow · GitHub
Skip to content
[e2e] Re-enable concurrent execution of the e2e suite - #3658
Merged
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Aug 20, 2026
Merged
[e2e] Re-enable concurrent execution of the e2e suite#3658
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Conversation
@alangenfeld
@alangenfeldalangenfeld
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Collaborator
Summary & Motivation
Serial execution was the dominant wall-clock cost per matrix entry (~22 of 24 minutes on the Vercel lanes), and the blockers that forced it in 78048e0 are fixed.
  • Per-test harness state (name, tracked runs, skip) binds through AsyncLocalStorage instead of module globals, so a failing test dumps its own diagnostics and a mid-body conformance gate skips the right test. The suite's test is a createTaskCollector wrapper that delegates to the enclosing suite's collector, which is both how the binding gets a direct call stack to the handler and how describe.concurrent inheritance reaches each test.
  • Each lane ends with a load summary — test and CLI child counts, peak concurrency of each, median test wall time, CLI share of summed test time, ten slowest tests — since under concurrency the question is where per-test latency went, not pass/fail.
  • maxConcurrency is tunable via WORKFLOW_E2E_MAX_CONCURRENCY, default unchanged at 5. Higher values starve CLI children on lanes that also host the app and Postgres.
  • RACE_WINNER_MAX_DURATION_MS goes 5s → 8s; queue latency under concurrency pushed observed winner durations to ~6.5s, still far below the 10s loser.
A test that genuinely cannot share a deployment can opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Test Plan
Added a unit test for interleaved per-test attribution; five consecutive full-matrix CI runs at maxConcurrency=5 were green (162/162) with Vercel-prod lanes at 5-7 min against 22-25 min serial and median test time unchanged, and a maxConcurrency=8 probe established 5 as the value safe on every lane type.
@changeset-bot
changeset-botBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

⚠️ No Changeset found
Latest commit: c35ea2c
Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.
This PR includes no changesets
When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types
Click here to learn what changesets are, and how to add one.
Click here if you're a maintainer who wants to add a changeset to this PR
@vercel
vercelBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
The latest updates on your projects. Learn more about Vercel for GitHub.
ProjectDeploymentActionsUpdated (UTC)
example-nextjs-workflow-turbopackReadyReadyPreviewAug 19, 2026 5:45pm
example-nextjs-workflow-webpackReadyReadyPreviewAug 19, 2026 5:45pm
example-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-astro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-express-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-fastify-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-hono-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nestjs-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nitro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nuxt-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-python-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-sveltekit-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-tanstack-start-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-vite-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workflow-docsReadyReadyPreview, v0Aug 19, 2026 5:45pm
workflow-swc-playgroundReadyReadyPreviewAug 19, 2026 5:45pm
workflow-tarballsReadyReadyPreviewAug 19, 2026 5:45pm
workflow-webReadyReadyPreviewAug 19, 2026 5:45pm
@alangenfeld
[e2e] Re-enable concurrent execution of the e2e suite
fc94f67
Serial execution has been the dominant wall-clock cost per matrix entry
since concurrency was disabled before conf (78048e0): ~128 tests at
~22 of 24 minutes on the Vercel lanes, and lately the slowest lane
cannot finish under its 30-minute job timeout on a slow runner day at
all. #2083 measured the concurrent suite at ~3x job wall-clock (4-5x on
the vitest phase) and identified what broke; its blockers are now fixed:
world-local writeExclusive is atomic (write-then-link), abort-fetch
tests are hermetic (#3618), the fibonacci tree fits the scheduler
(#3619), and source-map assertions are positive-only (#3620).
What this change adds is concurrency-safe per-test attribution. The
harness tracked runs and test names in module globals reset by a
beforeEach - under concurrency every test clobbered every other's
state, so a failing test dumped an unrelated sibling's diagnostics.
vitest's getCurrentTest() cannot substitute: it is a plain module
variable, wrong after any await. Instead an auto fixture - the one
place that receives the test's own context unambiguously - binds a
per-test state (name, tracked runs, the test's own skip) via
AsyncLocalStorage around each test body, and trackRun /
recordInfraEvent / requireFixture read it ambiently with no call-site
changes. The conformance gates skip through the bound state's skip, so
a mid-body requireFixture skips the right test. Sequential suites
(dev, agent, region) keep setupRunTracking's module-global fallback.
Full suite passes 137/137 concurrently against a local dev server in
under 2 minutes. A test that genuinely cannot share a deployment can
opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
[e2e] Bind per-test state via a task collector, not a fixture
f437a21
CI on the first concurrent run caught the fixture design failing in the
one place skip-correctness is load-bearing: vitest resolves fixtures in
a separate async context, so an AsyncLocalStorage store bound around
use() never reaches the test body. Every ambient read silently fell to
the module fallback - conformance-gated tests on the Python lane ran
instead of skipping and hard-failed on fixtures missing from the
manifest ('declared in e2e-conformance.json but not in the deployed
manifest').
The suite's test is now built with createTaskCollector, mirroring
vitest's own argument normalization, and the wrapper binds the state
around the handler call itself - a direct call stack, so propagation is
guaranteed rather than assumed. Validated by a collector-pattern ALS
experiment (per-test store after awaits, mid-body skip targets self,
each/skip/todo chains), a conformance-file simulation reproducing the
Python failure shape (fixture-missing tests now skip with the right
note), and a 137/137 concurrent local run.
Also raises RACE_WINNER_MAX_DURATION_MS 5s -> 8s: the winner takes 1s,
the loser 10s, and the bound only has to sit clearly below the loser -
concurrent-suite queue latency pushed observed winner durations to
~6.5s on loaded local-dev lanes (vite, first concurrent CI run), so 5s
flaked without catching anything 8s misses.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
alangenfeldforce-pushed
the
alangenfeld/e2e-concurrent-v2
branch
from
0556953 to
f437a21CompareAugust 19, 2026 15:05
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
📊 Workflow Benchmarks
commit c35ea2c · Wed, 19 Aug 2026 18:18:12 GMT · run logs
Backend: vercel · app: nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1380 (+381%) 🔻1517 🔴 (+30%) 🔻1557 🔴 (+16%) 🔻1624 🔴 (+3.9%)30
TTFSstream311 (+65%) 🔻1510 🔴 (+35%) 🔻1575 🔴 (+38%) 🔻1913 🔴 (+64%) 🔻30
TTFShook + stream1625 (+25%) 🔻1970 🔴 (+43%) 🔻2048 🔴 (+45%) 🔻2524 🔴 (+47%) 🔻30
Fan-out TTFSPromise.all(100 steps)1152 (+4.3%)2804 (+22%) 🔻3163 (+14%)3207 (+16%) 🔻10
Fan-out TTLSPromise.all(100 steps)5366 (+1.6%)8639 (+30%) 🔻10891 (+58%) 🔻19609 (+149%) 🔻10
STSO1020 steps (inline)117 (-11%)164 (-38%) 💚190 (-38%) 💚365 (-28%) 💚1019
WO1020 steps166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚1
CRTTfirst chunk (pooled)112 (+33%) 🔻186 (+46%) 🔻300 (-4.8%)436 (-30%) 💚28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)172 (+52%)176 (+7%)254 (+9%)457 (-2%)159 (-1%)10
size sweep (100/s, 160B-12KB)149 (+39%)165 (+7%)324 (+15%)518 (-24%)151 (-6%)10
replay gateway-gpt-5.4-nano-2000t (1x)126 (+13%)158 (-43%)240 (-67%)719 (-52%)353 (-41%)3
replay eve-gpt-5.6-sol-2000t (1x)135 (-14%)146 (+4%)186 (-3%)743 (+18%)523 (+7%)2
replay eve-gpt-5.6-sol-2000t (2x)138 (-73%)247 (-39%)355 (-87%)729 (-83%)569 (-6%)3
📈 STSO distribution vs main (inline / queue-hop histograms)
1020 steps (inline)
Cumulative STSO time: main 247209ms → this run 164994ms (Δ -82215ms, -33%)
100-150 ms █░░░░░░░░░░░░░░░░░░░░░░┃ main 6 this 474 +468
150-200 ms ██████████░░░░░░░░░░░░┃ main 205 this 464 +259
200-250 ms █┃██████████████████████ main 481 this 45 -436
250-300 ms ┃██████████ main 217 this 15 -202
300-350 ms ┃██ main 58 this 7 -51
350-400 ms ┃ main 20 this 8 -12
400-450 ms ┃ main 12 this 3 -9
450-500 ms ┃ main 9 this 1 -8
500-550 ms ┃ main 4 this 1 -3
550-600 ms ┃ main 2 this 0 -2
600-650 ms ┃ main 4 this 1 -3
850-900 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)
variant RTT 1ms→5s+ avg p50 p90 p99 n
control ······▂█▂···· 148.6 (+19%) 137 (+27%) 254 (+9%) 457 (-2%) 3000
sweep ······▃█▂▁··· 145.5 (+10%) 129 (+15%) 324 (+15%) 518 (-24%) 3000
gw 1x ·····▁▄█▁▁··· 132.2 (-25%) 116 (+5%) 240 (-67%) 719 (-52%) 5295
eve 1x ·····▁▅█▁▁··· 133.2 (+7%) 113 (+5%) 186 (-3%) 743 (+18%) 5186
eve 2x ·····▁▂█▃▁··· 173.2 (-55%) 139 (-19%) 355 (-87%) 729 (-83%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):
control █▄▃▂▄▂▄▁▁▂ 131–192ms
sweep ▇▃▂▁▃▃▄█▂▂ 124–181ms
gw 1x ▃█▄▂▂▄▁▂▃▄ 110–178ms
eve 1x ▂▃█▃▁▂▅▄▃█ 101–182ms
eve 2x █▁▂▂▁▄▅▅▄▅ 127–257ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):
sweep █▆▃▁▄▅█ 142–148ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):
control ▁▇▆▅▇▄█▄▆▄ 27–47ms
sweep ▁▆▆▆█▇▇▇▅▄ 31–61ms
gw 1x ▄▆▄▁▄█▃▄▃▅ 32–45ms
eve 1x ▄▆▆▇▁▄█▆▃█ 20–29ms
eve 2x █▁▃▂▃▅▃▁▄▅ 18–35ms

📜 Previous results (2)
c35ea2c
Wed, 19 Aug 2026 17:11:18 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1255 (+337%) 🔻1314 🔴 (+13%)1358 🔴 (+1.4%)1383 🔴 (-12%)30
TTFSstream241 (+28%) 🔻1334 🔴 (+20%) 🔻1399 🔴 (+23%) 🔻1681 🔴 (+44%) 🔻30
TTFShook + stream367 (-72%) 💚1748 🔴 (+27%) 🔻1778 🔴 (+26%) 🔻1952 🔴 (+14%)30
Fan-out TTFSPromise.all(100 steps)1392 (+26%) 🔻2936 (+28%) 🔻3132 (+13%)3151 (+14%)10
Fan-out TTLSPromise.all(100 steps)5963 (+13%)7084 (+6.4%)7237 (+5.1%)8436 (+7.0%)10
STSO1020 steps (inline)111 (-16%) 💚272 (+3.4%)320 (+5.3%)495 (-2.0%)1019
WO1020 steps245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)1
CRTTfirst chunk (pooled)98 (+17%) 🔻169 (+33%) 🔻398 (+26%) 🔻824 (+31%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)135 (+19%)434 (+163%)618 (+166%)1102 (+137%)617 (+287%)10
size sweep (100/s, 160B-12KB)149 (+39%)344 (+123%)540 (+92%)1044 (+54%)288 (+79%)10
replay gateway-gpt-5.4-nano-2000t (1x)127 (+13%)169 (-39%)290 (-60%)750 (-50%)524 (-13%)3
replay eve-gpt-5.6-sol-2000t (1x)206 (+32%)151 (+8%)246 (+28%)501 (-20%)380 (-22%)2
replay eve-gpt-5.6-sol-2000t (2x)153 (-70%)191 (-53%)315 (-89%)676 (-84%)444 (-27%)3
f437a21
Wed, 19 Aug 2026 15:26:31 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1341 (+367%) 🔻1586 🔴 (+36%) 🔻1690 🔴 (+26%) 🔻1885 🔴 (+21%) 🔻30
TTFSstream362 (+92%) 🔻1460 🔴 (+31%) 🔻1538 🔴 (+35%) 🔻1747 🔴 (+49%) 🔻30
TTFShook + stream1593 (+22%) 🔻1872 🔴 (+36%) 🔻1946 🔴 (+38%) 🔻2005 🔴 (+17%) 🔻30
Fan-out TTFSPromise.all(100 steps)1248 (+13%)2227 (-3.2%)2780 (+0.5%)3913 (+41%) 🔻10
Fan-out TTLSPromise.all(100 steps)5426 (+2.7%)11957 (+80%) 🔻22902 (+233%) 🔻24329 (+209%) 🔻10
STSO1020 steps (inline)111 (-16%) 💚164 (-38%) 💚187 (-38%) 💚285 (-44%) 💚1019
WO1020 steps159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚1
CRTTfirst chunk (pooled)96 (+14%)204 (+61%) 🔻341 (+8.3%)765 (+22%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)141 (+25%)184 (+12%)398 (+72%)735 (+58%)152 (-5%)10
size sweep (100/s, 160B-12KB)168 (+57%)155 (+1%)206 (-27%)408 (-40%)126 (-22%)10
replay gateway-gpt-5.4-nano-2000t (1x)316 (+182%)167 (-40%)225 (-69%)429 (-71%)293 (-51%)3
replay eve-gpt-5.6-sol-2000t (1x)161 (+3%)166 (+19%)217 (+13%)556 (-11%)270 (-45%)2
replay eve-gpt-5.6-sol-2000t (2x)162 (-68%)201 (-50%)264 (-91%)537 (-87%)177 (-71%)3
ℹ️ Metric definitions & methodology
Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.
The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.
The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.
Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.
Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)
Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles
Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041
🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600
All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).
Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
🧪 E2E Test Results
✅ All tests passed
⚠️ Flaky E2E Tests (passed on retry)
These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.
  • cancelRun via CLI - cancelling a running workflow (astro)
  • cancelRun via CLI - cancelling a running workflow (vite)
  • deploymentId: 'latest' is a no-op in non-Vercel worlds (nitro)
  • hookGetConflictThenStepParallelWorkflow - hook.getConflict() continuation step runs alongside other steps (nextjs-webpack)
🛠 Infra Events (absorbed by the harness)
Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.
  • cold-start-warmup · suite warmup (tanstack-start) · at 18:14:32Z · abandoned wrun_01M0DKNYTD6N9KWDRTEMB63XWM
  • run-pickup-stall · hookGetConflictWorkflow - hook.getConflict() resolves with the conflicting run when token is already registered (nuxt) · at 18:15:38Z · abandoned wrun_01M0DKRB1H8KCG69A7VFN9V5XN
E2E Test Summary
Summary
PassedFailedSkippedTotal
✅ ▲ Vercel Production347407384212
✅ 💻 Local Development381005584368
✅ 📦 Local Production381005584368
✅ 🐘 Local Postgres381005584368
✅ 🪟 Windows31200312
✅ 🌐 Cross-language Conformance90128137
✅ vercel-multi-region270027
Total152520254017792
Details by Category
✅ ▲ Vercel Production
AppPassedFailedSkipped
✅ astro-node128028
✅ astro-quickjs128028
✅ example-node128028
✅ example-quickjs128028
✅ express-node128028
✅ express-quickjs128028
✅ fastify-node128028
✅ fastify-quickjs128028
✅ hono-node128028
✅ hono-quickjs128028
✅ nest-node128028
✅ nest-quickjs128028
✅ nextjs-turbopack-node15303
✅ nextjs-turbopack-quickjs15303
✅ nextjs-webpack-node15303
✅ nextjs-webpack-quickjs15303
✅ nitro-node128028
✅ nitro-quickjs128028
✅ nuxt-node128028
✅ nuxt-quickjs128028
✅ python-node80148
✅ sveltekit-node14709
✅ sveltekit-quickjs14709
✅ tanstack-start-node128028
✅ tanstack-start-quickjs128028
✅ vite-node128028
✅ vite-quickjs128028
✅ 💻 Local Development
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 📦 Local Production
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🐘 Local Postgres
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🪟 Windows
AppPassedFailedSkipped
✅ nextjs-turbopack-node15600
✅ nextjs-turbopack-quickjs15600
✅ 🌐 Cross-language Conformance
AppPassedFailedSkipped
✅ python90128
✅ vercel-multi-region
AppPassedFailedSkipped
✅ nextjs-turbopack2700
📋 View full workflow run
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
Sim World
Simulated world deterministic testing for races. Traces
🟠 world-sim scenario book — 1 fail of 41 total
fence=per-spec
scenariooutcomeeventsvirtreplayviolations
✅ smoke-no-stepscompleted30msok0
✅ smoke-one-stepcompleted60msok0
✅ hook-at-step-startedcompleted120msok0
✅ hook-at-step-completedcompleted120msok0
✅ hook-at-hook-createdcompleted120msok0
✅ deadline-hook-winscompleted71.0hok0
✅ deadline-expirescompleted71.0hok0
✅ long-sleepcompleted1130.0dok0
✅ hook-never-arrivesstalled30msskipped0
✅ step-retries-twicecompleted102.0sok0
✅ parallel-stepscompleted90msok0
✅ hook-on-execution-statecompleted120msok0
✅ peek-hook-before-branchcompleted120msok0
✅ peek-hook-after-branchcompleted120msok0
✅ peek-hook-at-registrationcompleted120msok0
✅ race-hook-before-probecompleted120msok0
✅ race-hook-after-probecompleted120msok0
✅ race-duplicate-deliverycompleted130msok0
✅ attr-hook-before-stepcompleted110msok0
✅ attr-hook-after-stepcompleted110msok0
✅ attr-from-step-bodycompleted130msok0
✅ fork-hook-after-timeoutcompleted141.0mok0
✅ fork-hook-before-timeoutcompleted141.0mok0
✅ count-hook-after-timeoutcompleted171.0mok0
✅ count-hook-before-timeoutcompleted201.0mok0
✅ stale-read-step-count-forkcompleted201.0mok0
✅ stale-read-equal-step-countscompleted141.0mok0
✅ step-vs-step-forkcompleted120msok0
✅ step-vs-step-fork-fencedcompleted120msok0
✅ fence-catches-benign-directioncompleted125msok0
✅ in-flight-before-decisioncompleted171.0mok0
❌ in-flight-before-decision-countedcompleted171.0mok0
✅ in-flight-after-decisioncompleted192.0mok0
✅ stale-read-step-count-fork-fencedcompleted201.0mok0
✅ fork-hook-winscompleted131.0mok0
✅ fork-timeout-winscompleted131.0mok0
✅ unclaimed-payload-under-forkcompleted171.0mok0
✅ claimed-payload-under-forkcompleted171.0mok0
✅ writers-independent-step-bodiescompleted120msok0
✅ writers-scripted-tempocompleted120msok0
✅ cancel-mid-stepcancelled70msskipped0
Full trace: world-sim.txt
@VaguelySeriousVaguelySerious
mentioned this pull request
Aug 19, 2026
 Merged

@alangenfeld
[e2e] Delegate to the suite collector so describe.concurrent applies
ef36a7e
The previous commit's collector called getCurrentSuite().task() directly,
which skips the suite collector's Object.assign({}, suiteOptions,
options) - the merge that carries describe.concurrent onto each test. The
suite therefore still ran sequentially, and CI proved it: every Vercel
Prod lane matched the serial baseline minute-for-minute
(nextjs-turbopack quickjs 25m vs 25m), so a fully green run said nothing
about concurrency. The same slowdown showed up locally as a full-suite
run taking 6m15s where the earlier concurrent one took 1m58s, which I
had misread as server slowness.
The collector now wraps the handler and hands the call to
getCurrentSuite().test.fn - exactly what vitest's own top-level test
does - so suite-option inheritance is vitest's code path again, while
the wrapper still owns the handler call stack that makes the
AsyncLocalStorage binding reach the test body.
Verified with an experiment asserting both properties at once: four
300ms tests plus a 600ms observer finish in 602ms (1800ms if
sequential), max-in-flight > 1, and each test still reads its own store
after awaits.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
vercel[bot]
vercelBot
reviewed
Aug 19, 2026
View reviewed changes
Comment threadpackages/core/e2e/e2e.test.ts
@alangenfeld
[e2e] Actually call the load-observability hooks
c35ea2c
The previous commit added the counters and the summary but left the call
sites out of the handler wrapper and afterAll, so nothing was collected
and no summary printed. Caught by review.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@github-actions
github-actionsBot
commented
Aug 20, 2026

Copy link


Copy Markdown

Contributor
No backport to stable for 8c29b85 (AI decision).
This is test-harness enhancement work, not a fix: it re-enables describe.concurrent for the e2e suite to cut CI wall clock, adds a new load-summary reporting facility, a new WORKFLOW_E2E_MAX_CONCURRENCY config knob, and reworks per-test state onto AsyncLocalStorage. Nothing here repairs broken or flaky behavior on stable — it trades a known-serial, known-green suite for a concurrency model whose safe value was tuned against main's lanes, which is exactly the kind of risk a maintenance branch is avoiding. The RACE_WINNER_MAX_DURATION_MS 5s→8s bump is the only flake-adjacent piece, and it is only needed because of the concurrency this commit introduces.
To override, re-run the Backport to stable workflow manually via workflow_dispatch and paste this commit SHA into the ref input:
8c29b85322dc2279b6274391cd14a1483eb969c3

Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment

Reviewers

@vercelvercel[bot]vercel[bot] left review comments
@TooTallNateTooTallNateTooTallNate approved these changes

Assignees


No one assigned


Labels


None yet


Projects


None yet


Milestone


No milestone


Development

Successfully merging this pull request may close these issues.

3 participants

@alangenfeld@TooTallNate@VaguelySerious

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Force GitHub README to respect dark mode
(function() {
var style = document.createElement('style');
style.textContent = '
.markdown-body {
color-scheme: dark light;
}
.markdown-body pre { background: #161b22 !important; }
.markdown-body code { background: rgba(110, 118, 129, 0.4) !important; }
.markdown-body table th, .markdown-body table td { border-color: #30363d !important; }
.markdown-body img { background: #0d1117; }
.markdown-body blockquote { border-left-color: #8b949e; }
.markdown-body hr { border-color: #30363d; }
';
document.head.appendChild(style);
})();
}
} catch(__e) { console.warn('[Userscript:GitHub Dark Mode README Fix]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '
[e2e] Re-enable concurrent execution of the e2e suite by alangenfeld · Pull Request #3658 · vercel/workflow · GitHub
Skip to content
[e2e] Re-enable concurrent execution of the e2e suite - #3658
Merged
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Aug 20, 2026
Merged
[e2e] Re-enable concurrent execution of the e2e suite#3658
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Conversation
@alangenfeld
@alangenfeldalangenfeld
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Collaborator
Summary & Motivation
Serial execution was the dominant wall-clock cost per matrix entry (~22 of 24 minutes on the Vercel lanes), and the blockers that forced it in 78048e0 are fixed.
  • Per-test harness state (name, tracked runs, skip) binds through AsyncLocalStorage instead of module globals, so a failing test dumps its own diagnostics and a mid-body conformance gate skips the right test. The suite's test is a createTaskCollector wrapper that delegates to the enclosing suite's collector, which is both how the binding gets a direct call stack to the handler and how describe.concurrent inheritance reaches each test.
  • Each lane ends with a load summary — test and CLI child counts, peak concurrency of each, median test wall time, CLI share of summed test time, ten slowest tests — since under concurrency the question is where per-test latency went, not pass/fail.
  • maxConcurrency is tunable via WORKFLOW_E2E_MAX_CONCURRENCY, default unchanged at 5. Higher values starve CLI children on lanes that also host the app and Postgres.
  • RACE_WINNER_MAX_DURATION_MS goes 5s → 8s; queue latency under concurrency pushed observed winner durations to ~6.5s, still far below the 10s loser.
A test that genuinely cannot share a deployment can opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Test Plan
Added a unit test for interleaved per-test attribution; five consecutive full-matrix CI runs at maxConcurrency=5 were green (162/162) with Vercel-prod lanes at 5-7 min against 22-25 min serial and median test time unchanged, and a maxConcurrency=8 probe established 5 as the value safe on every lane type.
@changeset-bot
changeset-botBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

⚠️ No Changeset found
Latest commit: c35ea2c
Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.
This PR includes no changesets
When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types
Click here to learn what changesets are, and how to add one.
Click here if you're a maintainer who wants to add a changeset to this PR
@vercel
vercelBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
The latest updates on your projects. Learn more about Vercel for GitHub.
ProjectDeploymentActionsUpdated (UTC)
example-nextjs-workflow-turbopackReadyReadyPreviewAug 19, 2026 5:45pm
example-nextjs-workflow-webpackReadyReadyPreviewAug 19, 2026 5:45pm
example-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-astro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-express-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-fastify-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-hono-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nestjs-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nitro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nuxt-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-python-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-sveltekit-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-tanstack-start-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-vite-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workflow-docsReadyReadyPreview, v0Aug 19, 2026 5:45pm
workflow-swc-playgroundReadyReadyPreviewAug 19, 2026 5:45pm
workflow-tarballsReadyReadyPreviewAug 19, 2026 5:45pm
workflow-webReadyReadyPreviewAug 19, 2026 5:45pm
@alangenfeld
[e2e] Re-enable concurrent execution of the e2e suite
fc94f67
Serial execution has been the dominant wall-clock cost per matrix entry
since concurrency was disabled before conf (78048e0): ~128 tests at
~22 of 24 minutes on the Vercel lanes, and lately the slowest lane
cannot finish under its 30-minute job timeout on a slow runner day at
all. #2083 measured the concurrent suite at ~3x job wall-clock (4-5x on
the vitest phase) and identified what broke; its blockers are now fixed:
world-local writeExclusive is atomic (write-then-link), abort-fetch
tests are hermetic (#3618), the fibonacci tree fits the scheduler
(#3619), and source-map assertions are positive-only (#3620).
What this change adds is concurrency-safe per-test attribution. The
harness tracked runs and test names in module globals reset by a
beforeEach - under concurrency every test clobbered every other's
state, so a failing test dumped an unrelated sibling's diagnostics.
vitest's getCurrentTest() cannot substitute: it is a plain module
variable, wrong after any await. Instead an auto fixture - the one
place that receives the test's own context unambiguously - binds a
per-test state (name, tracked runs, the test's own skip) via
AsyncLocalStorage around each test body, and trackRun /
recordInfraEvent / requireFixture read it ambiently with no call-site
changes. The conformance gates skip through the bound state's skip, so
a mid-body requireFixture skips the right test. Sequential suites
(dev, agent, region) keep setupRunTracking's module-global fallback.
Full suite passes 137/137 concurrently against a local dev server in
under 2 minutes. A test that genuinely cannot share a deployment can
opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
[e2e] Bind per-test state via a task collector, not a fixture
f437a21
CI on the first concurrent run caught the fixture design failing in the
one place skip-correctness is load-bearing: vitest resolves fixtures in
a separate async context, so an AsyncLocalStorage store bound around
use() never reaches the test body. Every ambient read silently fell to
the module fallback - conformance-gated tests on the Python lane ran
instead of skipping and hard-failed on fixtures missing from the
manifest ('declared in e2e-conformance.json but not in the deployed
manifest').
The suite's test is now built with createTaskCollector, mirroring
vitest's own argument normalization, and the wrapper binds the state
around the handler call itself - a direct call stack, so propagation is
guaranteed rather than assumed. Validated by a collector-pattern ALS
experiment (per-test store after awaits, mid-body skip targets self,
each/skip/todo chains), a conformance-file simulation reproducing the
Python failure shape (fixture-missing tests now skip with the right
note), and a 137/137 concurrent local run.
Also raises RACE_WINNER_MAX_DURATION_MS 5s -> 8s: the winner takes 1s,
the loser 10s, and the bound only has to sit clearly below the loser -
concurrent-suite queue latency pushed observed winner durations to
~6.5s on loaded local-dev lanes (vite, first concurrent CI run), so 5s
flaked without catching anything 8s misses.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
alangenfeldforce-pushed
the
alangenfeld/e2e-concurrent-v2
branch
from
0556953 to
f437a21CompareAugust 19, 2026 15:05
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
📊 Workflow Benchmarks
commit c35ea2c · Wed, 19 Aug 2026 18:18:12 GMT · run logs
Backend: vercel · app: nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1380 (+381%) 🔻1517 🔴 (+30%) 🔻1557 🔴 (+16%) 🔻1624 🔴 (+3.9%)30
TTFSstream311 (+65%) 🔻1510 🔴 (+35%) 🔻1575 🔴 (+38%) 🔻1913 🔴 (+64%) 🔻30
TTFShook + stream1625 (+25%) 🔻1970 🔴 (+43%) 🔻2048 🔴 (+45%) 🔻2524 🔴 (+47%) 🔻30
Fan-out TTFSPromise.all(100 steps)1152 (+4.3%)2804 (+22%) 🔻3163 (+14%)3207 (+16%) 🔻10
Fan-out TTLSPromise.all(100 steps)5366 (+1.6%)8639 (+30%) 🔻10891 (+58%) 🔻19609 (+149%) 🔻10
STSO1020 steps (inline)117 (-11%)164 (-38%) 💚190 (-38%) 💚365 (-28%) 💚1019
WO1020 steps166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚1
CRTTfirst chunk (pooled)112 (+33%) 🔻186 (+46%) 🔻300 (-4.8%)436 (-30%) 💚28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)172 (+52%)176 (+7%)254 (+9%)457 (-2%)159 (-1%)10
size sweep (100/s, 160B-12KB)149 (+39%)165 (+7%)324 (+15%)518 (-24%)151 (-6%)10
replay gateway-gpt-5.4-nano-2000t (1x)126 (+13%)158 (-43%)240 (-67%)719 (-52%)353 (-41%)3
replay eve-gpt-5.6-sol-2000t (1x)135 (-14%)146 (+4%)186 (-3%)743 (+18%)523 (+7%)2
replay eve-gpt-5.6-sol-2000t (2x)138 (-73%)247 (-39%)355 (-87%)729 (-83%)569 (-6%)3
📈 STSO distribution vs main (inline / queue-hop histograms)
1020 steps (inline)
Cumulative STSO time: main 247209ms → this run 164994ms (Δ -82215ms, -33%)
100-150 ms █░░░░░░░░░░░░░░░░░░░░░░┃ main 6 this 474 +468
150-200 ms ██████████░░░░░░░░░░░░┃ main 205 this 464 +259
200-250 ms █┃██████████████████████ main 481 this 45 -436
250-300 ms ┃██████████ main 217 this 15 -202
300-350 ms ┃██ main 58 this 7 -51
350-400 ms ┃ main 20 this 8 -12
400-450 ms ┃ main 12 this 3 -9
450-500 ms ┃ main 9 this 1 -8
500-550 ms ┃ main 4 this 1 -3
550-600 ms ┃ main 2 this 0 -2
600-650 ms ┃ main 4 this 1 -3
850-900 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)
variant RTT 1ms→5s+ avg p50 p90 p99 n
control ······▂█▂···· 148.6 (+19%) 137 (+27%) 254 (+9%) 457 (-2%) 3000
sweep ······▃█▂▁··· 145.5 (+10%) 129 (+15%) 324 (+15%) 518 (-24%) 3000
gw 1x ·····▁▄█▁▁··· 132.2 (-25%) 116 (+5%) 240 (-67%) 719 (-52%) 5295
eve 1x ·····▁▅█▁▁··· 133.2 (+7%) 113 (+5%) 186 (-3%) 743 (+18%) 5186
eve 2x ·····▁▂█▃▁··· 173.2 (-55%) 139 (-19%) 355 (-87%) 729 (-83%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):
control █▄▃▂▄▂▄▁▁▂ 131–192ms
sweep ▇▃▂▁▃▃▄█▂▂ 124–181ms
gw 1x ▃█▄▂▂▄▁▂▃▄ 110–178ms
eve 1x ▂▃█▃▁▂▅▄▃█ 101–182ms
eve 2x █▁▂▂▁▄▅▅▄▅ 127–257ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):
sweep █▆▃▁▄▅█ 142–148ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):
control ▁▇▆▅▇▄█▄▆▄ 27–47ms
sweep ▁▆▆▆█▇▇▇▅▄ 31–61ms
gw 1x ▄▆▄▁▄█▃▄▃▅ 32–45ms
eve 1x ▄▆▆▇▁▄█▆▃█ 20–29ms
eve 2x █▁▃▂▃▅▃▁▄▅ 18–35ms

📜 Previous results (2)
c35ea2c
Wed, 19 Aug 2026 17:11:18 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1255 (+337%) 🔻1314 🔴 (+13%)1358 🔴 (+1.4%)1383 🔴 (-12%)30
TTFSstream241 (+28%) 🔻1334 🔴 (+20%) 🔻1399 🔴 (+23%) 🔻1681 🔴 (+44%) 🔻30
TTFShook + stream367 (-72%) 💚1748 🔴 (+27%) 🔻1778 🔴 (+26%) 🔻1952 🔴 (+14%)30
Fan-out TTFSPromise.all(100 steps)1392 (+26%) 🔻2936 (+28%) 🔻3132 (+13%)3151 (+14%)10
Fan-out TTLSPromise.all(100 steps)5963 (+13%)7084 (+6.4%)7237 (+5.1%)8436 (+7.0%)10
STSO1020 steps (inline)111 (-16%) 💚272 (+3.4%)320 (+5.3%)495 (-2.0%)1019
WO1020 steps245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)1
CRTTfirst chunk (pooled)98 (+17%) 🔻169 (+33%) 🔻398 (+26%) 🔻824 (+31%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)135 (+19%)434 (+163%)618 (+166%)1102 (+137%)617 (+287%)10
size sweep (100/s, 160B-12KB)149 (+39%)344 (+123%)540 (+92%)1044 (+54%)288 (+79%)10
replay gateway-gpt-5.4-nano-2000t (1x)127 (+13%)169 (-39%)290 (-60%)750 (-50%)524 (-13%)3
replay eve-gpt-5.6-sol-2000t (1x)206 (+32%)151 (+8%)246 (+28%)501 (-20%)380 (-22%)2
replay eve-gpt-5.6-sol-2000t (2x)153 (-70%)191 (-53%)315 (-89%)676 (-84%)444 (-27%)3
f437a21
Wed, 19 Aug 2026 15:26:31 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1341 (+367%) 🔻1586 🔴 (+36%) 🔻1690 🔴 (+26%) 🔻1885 🔴 (+21%) 🔻30
TTFSstream362 (+92%) 🔻1460 🔴 (+31%) 🔻1538 🔴 (+35%) 🔻1747 🔴 (+49%) 🔻30
TTFShook + stream1593 (+22%) 🔻1872 🔴 (+36%) 🔻1946 🔴 (+38%) 🔻2005 🔴 (+17%) 🔻30
Fan-out TTFSPromise.all(100 steps)1248 (+13%)2227 (-3.2%)2780 (+0.5%)3913 (+41%) 🔻10
Fan-out TTLSPromise.all(100 steps)5426 (+2.7%)11957 (+80%) 🔻22902 (+233%) 🔻24329 (+209%) 🔻10
STSO1020 steps (inline)111 (-16%) 💚164 (-38%) 💚187 (-38%) 💚285 (-44%) 💚1019
WO1020 steps159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚1
CRTTfirst chunk (pooled)96 (+14%)204 (+61%) 🔻341 (+8.3%)765 (+22%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)141 (+25%)184 (+12%)398 (+72%)735 (+58%)152 (-5%)10
size sweep (100/s, 160B-12KB)168 (+57%)155 (+1%)206 (-27%)408 (-40%)126 (-22%)10
replay gateway-gpt-5.4-nano-2000t (1x)316 (+182%)167 (-40%)225 (-69%)429 (-71%)293 (-51%)3
replay eve-gpt-5.6-sol-2000t (1x)161 (+3%)166 (+19%)217 (+13%)556 (-11%)270 (-45%)2
replay eve-gpt-5.6-sol-2000t (2x)162 (-68%)201 (-50%)264 (-91%)537 (-87%)177 (-71%)3
ℹ️ Metric definitions & methodology
Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.
The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.
The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.
Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.
Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)
Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles
Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041
🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600
All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).
Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
🧪 E2E Test Results
✅ All tests passed
⚠️ Flaky E2E Tests (passed on retry)
These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.
  • cancelRun via CLI - cancelling a running workflow (astro)
  • cancelRun via CLI - cancelling a running workflow (vite)
  • deploymentId: 'latest' is a no-op in non-Vercel worlds (nitro)
  • hookGetConflictThenStepParallelWorkflow - hook.getConflict() continuation step runs alongside other steps (nextjs-webpack)
🛠 Infra Events (absorbed by the harness)
Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.
  • cold-start-warmup · suite warmup (tanstack-start) · at 18:14:32Z · abandoned wrun_01M0DKNYTD6N9KWDRTEMB63XWM
  • run-pickup-stall · hookGetConflictWorkflow - hook.getConflict() resolves with the conflicting run when token is already registered (nuxt) · at 18:15:38Z · abandoned wrun_01M0DKRB1H8KCG69A7VFN9V5XN
E2E Test Summary
Summary
PassedFailedSkippedTotal
✅ ▲ Vercel Production347407384212
✅ 💻 Local Development381005584368
✅ 📦 Local Production381005584368
✅ 🐘 Local Postgres381005584368
✅ 🪟 Windows31200312
✅ 🌐 Cross-language Conformance90128137
✅ vercel-multi-region270027
Total152520254017792
Details by Category
✅ ▲ Vercel Production
AppPassedFailedSkipped
✅ astro-node128028
✅ astro-quickjs128028
✅ example-node128028
✅ example-quickjs128028
✅ express-node128028
✅ express-quickjs128028
✅ fastify-node128028
✅ fastify-quickjs128028
✅ hono-node128028
✅ hono-quickjs128028
✅ nest-node128028
✅ nest-quickjs128028
✅ nextjs-turbopack-node15303
✅ nextjs-turbopack-quickjs15303
✅ nextjs-webpack-node15303
✅ nextjs-webpack-quickjs15303
✅ nitro-node128028
✅ nitro-quickjs128028
✅ nuxt-node128028
✅ nuxt-quickjs128028
✅ python-node80148
✅ sveltekit-node14709
✅ sveltekit-quickjs14709
✅ tanstack-start-node128028
✅ tanstack-start-quickjs128028
✅ vite-node128028
✅ vite-quickjs128028
✅ 💻 Local Development
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 📦 Local Production
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🐘 Local Postgres
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🪟 Windows
AppPassedFailedSkipped
✅ nextjs-turbopack-node15600
✅ nextjs-turbopack-quickjs15600
✅ 🌐 Cross-language Conformance
AppPassedFailedSkipped
✅ python90128
✅ vercel-multi-region
AppPassedFailedSkipped
✅ nextjs-turbopack2700
📋 View full workflow run
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
Sim World
Simulated world deterministic testing for races. Traces
🟠 world-sim scenario book — 1 fail of 41 total
fence=per-spec
scenariooutcomeeventsvirtreplayviolations
✅ smoke-no-stepscompleted30msok0
✅ smoke-one-stepcompleted60msok0
✅ hook-at-step-startedcompleted120msok0
✅ hook-at-step-completedcompleted120msok0
✅ hook-at-hook-createdcompleted120msok0
✅ deadline-hook-winscompleted71.0hok0
✅ deadline-expirescompleted71.0hok0
✅ long-sleepcompleted1130.0dok0
✅ hook-never-arrivesstalled30msskipped0
✅ step-retries-twicecompleted102.0sok0
✅ parallel-stepscompleted90msok0
✅ hook-on-execution-statecompleted120msok0
✅ peek-hook-before-branchcompleted120msok0
✅ peek-hook-after-branchcompleted120msok0
✅ peek-hook-at-registrationcompleted120msok0
✅ race-hook-before-probecompleted120msok0
✅ race-hook-after-probecompleted120msok0
✅ race-duplicate-deliverycompleted130msok0
✅ attr-hook-before-stepcompleted110msok0
✅ attr-hook-after-stepcompleted110msok0
✅ attr-from-step-bodycompleted130msok0
✅ fork-hook-after-timeoutcompleted141.0mok0
✅ fork-hook-before-timeoutcompleted141.0mok0
✅ count-hook-after-timeoutcompleted171.0mok0
✅ count-hook-before-timeoutcompleted201.0mok0
✅ stale-read-step-count-forkcompleted201.0mok0
✅ stale-read-equal-step-countscompleted141.0mok0
✅ step-vs-step-forkcompleted120msok0
✅ step-vs-step-fork-fencedcompleted120msok0
✅ fence-catches-benign-directioncompleted125msok0
✅ in-flight-before-decisioncompleted171.0mok0
❌ in-flight-before-decision-countedcompleted171.0mok0
✅ in-flight-after-decisioncompleted192.0mok0
✅ stale-read-step-count-fork-fencedcompleted201.0mok0
✅ fork-hook-winscompleted131.0mok0
✅ fork-timeout-winscompleted131.0mok0
✅ unclaimed-payload-under-forkcompleted171.0mok0
✅ claimed-payload-under-forkcompleted171.0mok0
✅ writers-independent-step-bodiescompleted120msok0
✅ writers-scripted-tempocompleted120msok0
✅ cancel-mid-stepcancelled70msskipped0
Full trace: world-sim.txt
@VaguelySeriousVaguelySerious
mentioned this pull request
Aug 19, 2026
 Merged

@alangenfeld
[e2e] Delegate to the suite collector so describe.concurrent applies
ef36a7e
The previous commit's collector called getCurrentSuite().task() directly,
which skips the suite collector's Object.assign({}, suiteOptions,
options) - the merge that carries describe.concurrent onto each test. The
suite therefore still ran sequentially, and CI proved it: every Vercel
Prod lane matched the serial baseline minute-for-minute
(nextjs-turbopack quickjs 25m vs 25m), so a fully green run said nothing
about concurrency. The same slowdown showed up locally as a full-suite
run taking 6m15s where the earlier concurrent one took 1m58s, which I
had misread as server slowness.
The collector now wraps the handler and hands the call to
getCurrentSuite().test.fn - exactly what vitest's own top-level test
does - so suite-option inheritance is vitest's code path again, while
the wrapper still owns the handler call stack that makes the
AsyncLocalStorage binding reach the test body.
Verified with an experiment asserting both properties at once: four
300ms tests plus a 600ms observer finish in 602ms (1800ms if
sequential), max-in-flight > 1, and each test still reads its own store
after awaits.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
vercel[bot]
vercelBot
reviewed
Aug 19, 2026
View reviewed changes
Comment threadpackages/core/e2e/e2e.test.ts
@alangenfeld
[e2e] Actually call the load-observability hooks
c35ea2c
The previous commit added the counters and the summary but left the call
sites out of the handler wrapper and afterAll, so nothing was collected
and no summary printed. Caught by review.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@github-actions
github-actionsBot
commented
Aug 20, 2026

Copy link


Copy Markdown

Contributor
No backport to stable for 8c29b85 (AI decision).
This is test-harness enhancement work, not a fix: it re-enables describe.concurrent for the e2e suite to cut CI wall clock, adds a new load-summary reporting facility, a new WORKFLOW_E2E_MAX_CONCURRENCY config knob, and reworks per-test state onto AsyncLocalStorage. Nothing here repairs broken or flaky behavior on stable — it trades a known-serial, known-green suite for a concurrency model whose safe value was tuned against main's lanes, which is exactly the kind of risk a maintenance branch is avoiding. The RACE_WINNER_MAX_DURATION_MS 5s→8s bump is the only flake-adjacent piece, and it is only needed because of the concurrency this commit introduces.
To override, re-run the Backport to stable workflow manually via workflow_dispatch and paste this commit SHA into the ref input:
8c29b85322dc2279b6274391cd14a1483eb969c3

Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment

Reviewers

@vercelvercel[bot]vercel[bot] left review comments
@TooTallNateTooTallNateTooTallNate approved these changes

Assignees


No one assigned


Labels


None yet


Projects


None yet


Milestone


No milestone


Development

Successfully merging this pull request may close these issues.

3 participants

@alangenfeld@TooTallNate@VaguelySerious

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Highlight search terms from Google/DuckDuckGo/Bing referrer
(function() {
var ref = document.referrer;
var terms = [];
if (ref.includes('google.com') || ref.includes('duckduckgo.com') || ref.includes('bing.com')) {
var url = new URL(ref);
var q = url.searchParams.get('q') || url.searchParams.get('p');
if (q) {
terms = q.split(/\s+/).filter(function(t) { return t.length > 2; });
}
}
if (terms.length === 0) return;
var style = document.createElement('style');
style.textContent = '.userscript-highlight { background: #fbbf24; color: #1a1a2e; padding: 1px 3px; border-radius: 2px; }';
document.head.appendChild(style);
function highlight(node) {
if (node.nodeType === 3) { // text node
var text = node.textContent;
var found = false;
terms.forEach(function(term) {
var regex = new RegExp('(' + term.replace(/[.*+?^${}()|[\]\\]/g, '\\') + ')', 'gi');
if (regex.test(text)) {
found = true;
var frag = document.createDocumentFragment();
var parts = text.split(regex);
parts.forEach(function(part, i) {
if (i % 2 === 0) {
frag.appendChild(document.createTextNode(part));
} else {
var span = document.createElement('span');
span.className = 'userscript-highlight';
span.textContent = part;
frag.appendChild(span);
}
});
node.parentNode.replaceChild(frag, node);
}
});
} else if (node.nodeType === 1 && node.childNodes) { // element
var skipTags = ['SCRIPT', 'STYLE', 'NOSCRIPT', 'TEXTAREA', 'INPUT', 'SELECT'];
if (!skipTags.includes(node.tagName)) {
Array.from(node.childNodes).forEach(highlight);
}
}
}
highlight(document.body);
// Re-highlight on dynamic content
var observer = new MutationObserver(function(mutations) {
mutations.forEach(function(m) {
m.addedNodes.forEach(function(node) {
if (node.nodeType === 1 || node.nodeType === 3) highlight(node);
});
});
});
observer.observe(document.body, { childList: true, subtree: true });
})();
}
} catch(__e) { console.warn('[Userscript:Highlight Search Terms]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '
[e2e] Re-enable concurrent execution of the e2e suite by alangenfeld · Pull Request #3658 · vercel/workflow · GitHub
Skip to content
[e2e] Re-enable concurrent execution of the e2e suite - #3658
Merged
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Aug 20, 2026
Merged
[e2e] Re-enable concurrent execution of the e2e suite#3658
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Conversation
@alangenfeld
@alangenfeldalangenfeld
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Collaborator
Summary & Motivation
Serial execution was the dominant wall-clock cost per matrix entry (~22 of 24 minutes on the Vercel lanes), and the blockers that forced it in 78048e0 are fixed.
  • Per-test harness state (name, tracked runs, skip) binds through AsyncLocalStorage instead of module globals, so a failing test dumps its own diagnostics and a mid-body conformance gate skips the right test. The suite's test is a createTaskCollector wrapper that delegates to the enclosing suite's collector, which is both how the binding gets a direct call stack to the handler and how describe.concurrent inheritance reaches each test.
  • Each lane ends with a load summary — test and CLI child counts, peak concurrency of each, median test wall time, CLI share of summed test time, ten slowest tests — since under concurrency the question is where per-test latency went, not pass/fail.
  • maxConcurrency is tunable via WORKFLOW_E2E_MAX_CONCURRENCY, default unchanged at 5. Higher values starve CLI children on lanes that also host the app and Postgres.
  • RACE_WINNER_MAX_DURATION_MS goes 5s → 8s; queue latency under concurrency pushed observed winner durations to ~6.5s, still far below the 10s loser.
A test that genuinely cannot share a deployment can opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Test Plan
Added a unit test for interleaved per-test attribution; five consecutive full-matrix CI runs at maxConcurrency=5 were green (162/162) with Vercel-prod lanes at 5-7 min against 22-25 min serial and median test time unchanged, and a maxConcurrency=8 probe established 5 as the value safe on every lane type.
@changeset-bot
changeset-botBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

⚠️ No Changeset found
Latest commit: c35ea2c
Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.
This PR includes no changesets
When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types
Click here to learn what changesets are, and how to add one.
Click here if you're a maintainer who wants to add a changeset to this PR
@vercel
vercelBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
The latest updates on your projects. Learn more about Vercel for GitHub.
ProjectDeploymentActionsUpdated (UTC)
example-nextjs-workflow-turbopackReadyReadyPreviewAug 19, 2026 5:45pm
example-nextjs-workflow-webpackReadyReadyPreviewAug 19, 2026 5:45pm
example-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-astro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-express-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-fastify-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-hono-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nestjs-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nitro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nuxt-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-python-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-sveltekit-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-tanstack-start-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-vite-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workflow-docsReadyReadyPreview, v0Aug 19, 2026 5:45pm
workflow-swc-playgroundReadyReadyPreviewAug 19, 2026 5:45pm
workflow-tarballsReadyReadyPreviewAug 19, 2026 5:45pm
workflow-webReadyReadyPreviewAug 19, 2026 5:45pm
@alangenfeld
[e2e] Re-enable concurrent execution of the e2e suite
fc94f67
Serial execution has been the dominant wall-clock cost per matrix entry
since concurrency was disabled before conf (78048e0): ~128 tests at
~22 of 24 minutes on the Vercel lanes, and lately the slowest lane
cannot finish under its 30-minute job timeout on a slow runner day at
all. #2083 measured the concurrent suite at ~3x job wall-clock (4-5x on
the vitest phase) and identified what broke; its blockers are now fixed:
world-local writeExclusive is atomic (write-then-link), abort-fetch
tests are hermetic (#3618), the fibonacci tree fits the scheduler
(#3619), and source-map assertions are positive-only (#3620).
What this change adds is concurrency-safe per-test attribution. The
harness tracked runs and test names in module globals reset by a
beforeEach - under concurrency every test clobbered every other's
state, so a failing test dumped an unrelated sibling's diagnostics.
vitest's getCurrentTest() cannot substitute: it is a plain module
variable, wrong after any await. Instead an auto fixture - the one
place that receives the test's own context unambiguously - binds a
per-test state (name, tracked runs, the test's own skip) via
AsyncLocalStorage around each test body, and trackRun /
recordInfraEvent / requireFixture read it ambiently with no call-site
changes. The conformance gates skip through the bound state's skip, so
a mid-body requireFixture skips the right test. Sequential suites
(dev, agent, region) keep setupRunTracking's module-global fallback.
Full suite passes 137/137 concurrently against a local dev server in
under 2 minutes. A test that genuinely cannot share a deployment can
opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
[e2e] Bind per-test state via a task collector, not a fixture
f437a21
CI on the first concurrent run caught the fixture design failing in the
one place skip-correctness is load-bearing: vitest resolves fixtures in
a separate async context, so an AsyncLocalStorage store bound around
use() never reaches the test body. Every ambient read silently fell to
the module fallback - conformance-gated tests on the Python lane ran
instead of skipping and hard-failed on fixtures missing from the
manifest ('declared in e2e-conformance.json but not in the deployed
manifest').
The suite's test is now built with createTaskCollector, mirroring
vitest's own argument normalization, and the wrapper binds the state
around the handler call itself - a direct call stack, so propagation is
guaranteed rather than assumed. Validated by a collector-pattern ALS
experiment (per-test store after awaits, mid-body skip targets self,
each/skip/todo chains), a conformance-file simulation reproducing the
Python failure shape (fixture-missing tests now skip with the right
note), and a 137/137 concurrent local run.
Also raises RACE_WINNER_MAX_DURATION_MS 5s -> 8s: the winner takes 1s,
the loser 10s, and the bound only has to sit clearly below the loser -
concurrent-suite queue latency pushed observed winner durations to
~6.5s on loaded local-dev lanes (vite, first concurrent CI run), so 5s
flaked without catching anything 8s misses.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
alangenfeldforce-pushed
the
alangenfeld/e2e-concurrent-v2
branch
from
0556953 to
f437a21CompareAugust 19, 2026 15:05
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
📊 Workflow Benchmarks
commit c35ea2c · Wed, 19 Aug 2026 18:18:12 GMT · run logs
Backend: vercel · app: nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1380 (+381%) 🔻1517 🔴 (+30%) 🔻1557 🔴 (+16%) 🔻1624 🔴 (+3.9%)30
TTFSstream311 (+65%) 🔻1510 🔴 (+35%) 🔻1575 🔴 (+38%) 🔻1913 🔴 (+64%) 🔻30
TTFShook + stream1625 (+25%) 🔻1970 🔴 (+43%) 🔻2048 🔴 (+45%) 🔻2524 🔴 (+47%) 🔻30
Fan-out TTFSPromise.all(100 steps)1152 (+4.3%)2804 (+22%) 🔻3163 (+14%)3207 (+16%) 🔻10
Fan-out TTLSPromise.all(100 steps)5366 (+1.6%)8639 (+30%) 🔻10891 (+58%) 🔻19609 (+149%) 🔻10
STSO1020 steps (inline)117 (-11%)164 (-38%) 💚190 (-38%) 💚365 (-28%) 💚1019
WO1020 steps166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚1
CRTTfirst chunk (pooled)112 (+33%) 🔻186 (+46%) 🔻300 (-4.8%)436 (-30%) 💚28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)172 (+52%)176 (+7%)254 (+9%)457 (-2%)159 (-1%)10
size sweep (100/s, 160B-12KB)149 (+39%)165 (+7%)324 (+15%)518 (-24%)151 (-6%)10
replay gateway-gpt-5.4-nano-2000t (1x)126 (+13%)158 (-43%)240 (-67%)719 (-52%)353 (-41%)3
replay eve-gpt-5.6-sol-2000t (1x)135 (-14%)146 (+4%)186 (-3%)743 (+18%)523 (+7%)2
replay eve-gpt-5.6-sol-2000t (2x)138 (-73%)247 (-39%)355 (-87%)729 (-83%)569 (-6%)3
📈 STSO distribution vs main (inline / queue-hop histograms)
1020 steps (inline)
Cumulative STSO time: main 247209ms → this run 164994ms (Δ -82215ms, -33%)
100-150 ms █░░░░░░░░░░░░░░░░░░░░░░┃ main 6 this 474 +468
150-200 ms ██████████░░░░░░░░░░░░┃ main 205 this 464 +259
200-250 ms █┃██████████████████████ main 481 this 45 -436
250-300 ms ┃██████████ main 217 this 15 -202
300-350 ms ┃██ main 58 this 7 -51
350-400 ms ┃ main 20 this 8 -12
400-450 ms ┃ main 12 this 3 -9
450-500 ms ┃ main 9 this 1 -8
500-550 ms ┃ main 4 this 1 -3
550-600 ms ┃ main 2 this 0 -2
600-650 ms ┃ main 4 this 1 -3
850-900 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)
variant RTT 1ms→5s+ avg p50 p90 p99 n
control ······▂█▂···· 148.6 (+19%) 137 (+27%) 254 (+9%) 457 (-2%) 3000
sweep ······▃█▂▁··· 145.5 (+10%) 129 (+15%) 324 (+15%) 518 (-24%) 3000
gw 1x ·····▁▄█▁▁··· 132.2 (-25%) 116 (+5%) 240 (-67%) 719 (-52%) 5295
eve 1x ·····▁▅█▁▁··· 133.2 (+7%) 113 (+5%) 186 (-3%) 743 (+18%) 5186
eve 2x ·····▁▂█▃▁··· 173.2 (-55%) 139 (-19%) 355 (-87%) 729 (-83%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):
control █▄▃▂▄▂▄▁▁▂ 131–192ms
sweep ▇▃▂▁▃▃▄█▂▂ 124–181ms
gw 1x ▃█▄▂▂▄▁▂▃▄ 110–178ms
eve 1x ▂▃█▃▁▂▅▄▃█ 101–182ms
eve 2x █▁▂▂▁▄▅▅▄▅ 127–257ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):
sweep █▆▃▁▄▅█ 142–148ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):
control ▁▇▆▅▇▄█▄▆▄ 27–47ms
sweep ▁▆▆▆█▇▇▇▅▄ 31–61ms
gw 1x ▄▆▄▁▄█▃▄▃▅ 32–45ms
eve 1x ▄▆▆▇▁▄█▆▃█ 20–29ms
eve 2x █▁▃▂▃▅▃▁▄▅ 18–35ms

📜 Previous results (2)
c35ea2c
Wed, 19 Aug 2026 17:11:18 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1255 (+337%) 🔻1314 🔴 (+13%)1358 🔴 (+1.4%)1383 🔴 (-12%)30
TTFSstream241 (+28%) 🔻1334 🔴 (+20%) 🔻1399 🔴 (+23%) 🔻1681 🔴 (+44%) 🔻30
TTFShook + stream367 (-72%) 💚1748 🔴 (+27%) 🔻1778 🔴 (+26%) 🔻1952 🔴 (+14%)30
Fan-out TTFSPromise.all(100 steps)1392 (+26%) 🔻2936 (+28%) 🔻3132 (+13%)3151 (+14%)10
Fan-out TTLSPromise.all(100 steps)5963 (+13%)7084 (+6.4%)7237 (+5.1%)8436 (+7.0%)10
STSO1020 steps (inline)111 (-16%) 💚272 (+3.4%)320 (+5.3%)495 (-2.0%)1019
WO1020 steps245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)1
CRTTfirst chunk (pooled)98 (+17%) 🔻169 (+33%) 🔻398 (+26%) 🔻824 (+31%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)135 (+19%)434 (+163%)618 (+166%)1102 (+137%)617 (+287%)10
size sweep (100/s, 160B-12KB)149 (+39%)344 (+123%)540 (+92%)1044 (+54%)288 (+79%)10
replay gateway-gpt-5.4-nano-2000t (1x)127 (+13%)169 (-39%)290 (-60%)750 (-50%)524 (-13%)3
replay eve-gpt-5.6-sol-2000t (1x)206 (+32%)151 (+8%)246 (+28%)501 (-20%)380 (-22%)2
replay eve-gpt-5.6-sol-2000t (2x)153 (-70%)191 (-53%)315 (-89%)676 (-84%)444 (-27%)3
f437a21
Wed, 19 Aug 2026 15:26:31 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1341 (+367%) 🔻1586 🔴 (+36%) 🔻1690 🔴 (+26%) 🔻1885 🔴 (+21%) 🔻30
TTFSstream362 (+92%) 🔻1460 🔴 (+31%) 🔻1538 🔴 (+35%) 🔻1747 🔴 (+49%) 🔻30
TTFShook + stream1593 (+22%) 🔻1872 🔴 (+36%) 🔻1946 🔴 (+38%) 🔻2005 🔴 (+17%) 🔻30
Fan-out TTFSPromise.all(100 steps)1248 (+13%)2227 (-3.2%)2780 (+0.5%)3913 (+41%) 🔻10
Fan-out TTLSPromise.all(100 steps)5426 (+2.7%)11957 (+80%) 🔻22902 (+233%) 🔻24329 (+209%) 🔻10
STSO1020 steps (inline)111 (-16%) 💚164 (-38%) 💚187 (-38%) 💚285 (-44%) 💚1019
WO1020 steps159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚1
CRTTfirst chunk (pooled)96 (+14%)204 (+61%) 🔻341 (+8.3%)765 (+22%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)141 (+25%)184 (+12%)398 (+72%)735 (+58%)152 (-5%)10
size sweep (100/s, 160B-12KB)168 (+57%)155 (+1%)206 (-27%)408 (-40%)126 (-22%)10
replay gateway-gpt-5.4-nano-2000t (1x)316 (+182%)167 (-40%)225 (-69%)429 (-71%)293 (-51%)3
replay eve-gpt-5.6-sol-2000t (1x)161 (+3%)166 (+19%)217 (+13%)556 (-11%)270 (-45%)2
replay eve-gpt-5.6-sol-2000t (2x)162 (-68%)201 (-50%)264 (-91%)537 (-87%)177 (-71%)3
ℹ️ Metric definitions & methodology
Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.
The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.
The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.
Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.
Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)
Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles
Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041
🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600
All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).
Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
🧪 E2E Test Results
✅ All tests passed
⚠️ Flaky E2E Tests (passed on retry)
These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.
  • cancelRun via CLI - cancelling a running workflow (astro)
  • cancelRun via CLI - cancelling a running workflow (vite)
  • deploymentId: 'latest' is a no-op in non-Vercel worlds (nitro)
  • hookGetConflictThenStepParallelWorkflow - hook.getConflict() continuation step runs alongside other steps (nextjs-webpack)
🛠 Infra Events (absorbed by the harness)
Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.
  • cold-start-warmup · suite warmup (tanstack-start) · at 18:14:32Z · abandoned wrun_01M0DKNYTD6N9KWDRTEMB63XWM
  • run-pickup-stall · hookGetConflictWorkflow - hook.getConflict() resolves with the conflicting run when token is already registered (nuxt) · at 18:15:38Z · abandoned wrun_01M0DKRB1H8KCG69A7VFN9V5XN
E2E Test Summary
Summary
PassedFailedSkippedTotal
✅ ▲ Vercel Production347407384212
✅ 💻 Local Development381005584368
✅ 📦 Local Production381005584368
✅ 🐘 Local Postgres381005584368
✅ 🪟 Windows31200312
✅ 🌐 Cross-language Conformance90128137
✅ vercel-multi-region270027
Total152520254017792
Details by Category
✅ ▲ Vercel Production
AppPassedFailedSkipped
✅ astro-node128028
✅ astro-quickjs128028
✅ example-node128028
✅ example-quickjs128028
✅ express-node128028
✅ express-quickjs128028
✅ fastify-node128028
✅ fastify-quickjs128028
✅ hono-node128028
✅ hono-quickjs128028
✅ nest-node128028
✅ nest-quickjs128028
✅ nextjs-turbopack-node15303
✅ nextjs-turbopack-quickjs15303
✅ nextjs-webpack-node15303
✅ nextjs-webpack-quickjs15303
✅ nitro-node128028
✅ nitro-quickjs128028
✅ nuxt-node128028
✅ nuxt-quickjs128028
✅ python-node80148
✅ sveltekit-node14709
✅ sveltekit-quickjs14709
✅ tanstack-start-node128028
✅ tanstack-start-quickjs128028
✅ vite-node128028
✅ vite-quickjs128028
✅ 💻 Local Development
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 📦 Local Production
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🐘 Local Postgres
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🪟 Windows
AppPassedFailedSkipped
✅ nextjs-turbopack-node15600
✅ nextjs-turbopack-quickjs15600
✅ 🌐 Cross-language Conformance
AppPassedFailedSkipped
✅ python90128
✅ vercel-multi-region
AppPassedFailedSkipped
✅ nextjs-turbopack2700
📋 View full workflow run
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
Sim World
Simulated world deterministic testing for races. Traces
🟠 world-sim scenario book — 1 fail of 41 total
fence=per-spec
scenariooutcomeeventsvirtreplayviolations
✅ smoke-no-stepscompleted30msok0
✅ smoke-one-stepcompleted60msok0
✅ hook-at-step-startedcompleted120msok0
✅ hook-at-step-completedcompleted120msok0
✅ hook-at-hook-createdcompleted120msok0
✅ deadline-hook-winscompleted71.0hok0
✅ deadline-expirescompleted71.0hok0
✅ long-sleepcompleted1130.0dok0
✅ hook-never-arrivesstalled30msskipped0
✅ step-retries-twicecompleted102.0sok0
✅ parallel-stepscompleted90msok0
✅ hook-on-execution-statecompleted120msok0
✅ peek-hook-before-branchcompleted120msok0
✅ peek-hook-after-branchcompleted120msok0
✅ peek-hook-at-registrationcompleted120msok0
✅ race-hook-before-probecompleted120msok0
✅ race-hook-after-probecompleted120msok0
✅ race-duplicate-deliverycompleted130msok0
✅ attr-hook-before-stepcompleted110msok0
✅ attr-hook-after-stepcompleted110msok0
✅ attr-from-step-bodycompleted130msok0
✅ fork-hook-after-timeoutcompleted141.0mok0
✅ fork-hook-before-timeoutcompleted141.0mok0
✅ count-hook-after-timeoutcompleted171.0mok0
✅ count-hook-before-timeoutcompleted201.0mok0
✅ stale-read-step-count-forkcompleted201.0mok0
✅ stale-read-equal-step-countscompleted141.0mok0
✅ step-vs-step-forkcompleted120msok0
✅ step-vs-step-fork-fencedcompleted120msok0
✅ fence-catches-benign-directioncompleted125msok0
✅ in-flight-before-decisioncompleted171.0mok0
❌ in-flight-before-decision-countedcompleted171.0mok0
✅ in-flight-after-decisioncompleted192.0mok0
✅ stale-read-step-count-fork-fencedcompleted201.0mok0
✅ fork-hook-winscompleted131.0mok0
✅ fork-timeout-winscompleted131.0mok0
✅ unclaimed-payload-under-forkcompleted171.0mok0
✅ claimed-payload-under-forkcompleted171.0mok0
✅ writers-independent-step-bodiescompleted120msok0
✅ writers-scripted-tempocompleted120msok0
✅ cancel-mid-stepcancelled70msskipped0
Full trace: world-sim.txt
@VaguelySeriousVaguelySerious
mentioned this pull request
Aug 19, 2026
 Merged

@alangenfeld
[e2e] Delegate to the suite collector so describe.concurrent applies
ef36a7e
The previous commit's collector called getCurrentSuite().task() directly,
which skips the suite collector's Object.assign({}, suiteOptions,
options) - the merge that carries describe.concurrent onto each test. The
suite therefore still ran sequentially, and CI proved it: every Vercel
Prod lane matched the serial baseline minute-for-minute
(nextjs-turbopack quickjs 25m vs 25m), so a fully green run said nothing
about concurrency. The same slowdown showed up locally as a full-suite
run taking 6m15s where the earlier concurrent one took 1m58s, which I
had misread as server slowness.
The collector now wraps the handler and hands the call to
getCurrentSuite().test.fn - exactly what vitest's own top-level test
does - so suite-option inheritance is vitest's code path again, while
the wrapper still owns the handler call stack that makes the
AsyncLocalStorage binding reach the test body.
Verified with an experiment asserting both properties at once: four
300ms tests plus a 600ms observer finish in 602ms (1800ms if
sequential), max-in-flight > 1, and each test still reads its own store
after awaits.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
vercel[bot]
vercelBot
reviewed
Aug 19, 2026
View reviewed changes
Comment threadpackages/core/e2e/e2e.test.ts
@alangenfeld
[e2e] Actually call the load-observability hooks
c35ea2c
The previous commit added the counters and the summary but left the call
sites out of the handler wrapper and afterAll, so nothing was collected
and no summary printed. Caught by review.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@github-actions
github-actionsBot
commented
Aug 20, 2026

Copy link


Copy Markdown

Contributor
No backport to stable for 8c29b85 (AI decision).
This is test-harness enhancement work, not a fix: it re-enables describe.concurrent for the e2e suite to cut CI wall clock, adds a new load-summary reporting facility, a new WORKFLOW_E2E_MAX_CONCURRENCY config knob, and reworks per-test state onto AsyncLocalStorage. Nothing here repairs broken or flaky behavior on stable — it trades a known-serial, known-green suite for a concurrency model whose safe value was tuned against main's lanes, which is exactly the kind of risk a maintenance branch is avoiding. The RACE_WINNER_MAX_DURATION_MS 5s→8s bump is the only flake-adjacent piece, and it is only needed because of the concurrency this commit introduces.
To override, re-run the Backport to stable workflow manually via workflow_dispatch and paste this commit SHA into the ref input:
8c29b85322dc2279b6274391cd14a1483eb969c3

Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment

Reviewers

@vercelvercel[bot]vercel[bot] left review comments
@TooTallNateTooTallNateTooTallNate approved these changes

Assignees


No one assigned


Labels


None yet


Projects


None yet


Milestone


No milestone


Development

Successfully merging this pull request may close these issues.

3 participants

@alangenfeld@TooTallNate@VaguelySerious

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Strip utm_, fbclid, gclid, etc. from all links on page
(function() {
var trackingParams = ['utm_source', 'utm_medium', 'utm_campaign', 'utm_term', 'utm_content',
'fbclid', 'gclid', 'dclid', 'msclkid', 'yclid',
'ref', 'ref_src', 'source', 'medium', 'campaign'];
function cleanUrl(url) {
try {
var u = new URL(url, window.location.origin);
var changed = false;
trackingParams.forEach(function(p) {
if (u.searchParams.has(p)) {
u.searchParams.delete(p);
changed = true;
}
});
return changed ? u.toString() : url;
} catch (e) {
return url;
}
}
function cleanLinks() {
document.querySelectorAll('a[href]').forEach(function(a) {
var clean = cleanUrl(a.href);
if (clean !== a.href) a.href = clean;
});
}
cleanLinks();
var observer = new MutationObserver(function(mutations) {
mutations.forEach(function(m) {
m.addedNodes.forEach(function(node) {
if (node.nodeType === 1) {
if (node.tagName === 'A') cleanLinks();
node.querySelectorAll('a[href]').forEach(function(a) {
var clean = cleanUrl(a.href);
if (clean !== a.href) a.href = clean;
});
}
});
});
});
observer.observe(document.body, { childList: true, subtree: true });
})();
}
} catch(__e) { console.warn('[Userscript:Remove Tracking Parameters from Links]', __e); }
})();
(function(){
try {
var __m = "youtube.com";
var __re = new RegExp('^' + "youtube\\.com" + '
[e2e] Re-enable concurrent execution of the e2e suite by alangenfeld · Pull Request #3658 · vercel/workflow · GitHub
Skip to content
[e2e] Re-enable concurrent execution of the e2e suite - #3658
Merged
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Aug 20, 2026
Merged
[e2e] Re-enable concurrent execution of the e2e suite#3658
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Conversation
@alangenfeld
@alangenfeldalangenfeld
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Collaborator
Summary & Motivation
Serial execution was the dominant wall-clock cost per matrix entry (~22 of 24 minutes on the Vercel lanes), and the blockers that forced it in 78048e0 are fixed.
  • Per-test harness state (name, tracked runs, skip) binds through AsyncLocalStorage instead of module globals, so a failing test dumps its own diagnostics and a mid-body conformance gate skips the right test. The suite's test is a createTaskCollector wrapper that delegates to the enclosing suite's collector, which is both how the binding gets a direct call stack to the handler and how describe.concurrent inheritance reaches each test.
  • Each lane ends with a load summary — test and CLI child counts, peak concurrency of each, median test wall time, CLI share of summed test time, ten slowest tests — since under concurrency the question is where per-test latency went, not pass/fail.
  • maxConcurrency is tunable via WORKFLOW_E2E_MAX_CONCURRENCY, default unchanged at 5. Higher values starve CLI children on lanes that also host the app and Postgres.
  • RACE_WINNER_MAX_DURATION_MS goes 5s → 8s; queue latency under concurrency pushed observed winner durations to ~6.5s, still far below the 10s loser.
A test that genuinely cannot share a deployment can opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Test Plan
Added a unit test for interleaved per-test attribution; five consecutive full-matrix CI runs at maxConcurrency=5 were green (162/162) with Vercel-prod lanes at 5-7 min against 22-25 min serial and median test time unchanged, and a maxConcurrency=8 probe established 5 as the value safe on every lane type.
@changeset-bot
changeset-botBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

⚠️ No Changeset found
Latest commit: c35ea2c
Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.
This PR includes no changesets
When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types
Click here to learn what changesets are, and how to add one.
Click here if you're a maintainer who wants to add a changeset to this PR
@vercel
vercelBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
The latest updates on your projects. Learn more about Vercel for GitHub.
ProjectDeploymentActionsUpdated (UTC)
example-nextjs-workflow-turbopackReadyReadyPreviewAug 19, 2026 5:45pm
example-nextjs-workflow-webpackReadyReadyPreviewAug 19, 2026 5:45pm
example-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-astro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-express-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-fastify-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-hono-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nestjs-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nitro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nuxt-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-python-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-sveltekit-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-tanstack-start-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-vite-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workflow-docsReadyReadyPreview, v0Aug 19, 2026 5:45pm
workflow-swc-playgroundReadyReadyPreviewAug 19, 2026 5:45pm
workflow-tarballsReadyReadyPreviewAug 19, 2026 5:45pm
workflow-webReadyReadyPreviewAug 19, 2026 5:45pm
@alangenfeld
[e2e] Re-enable concurrent execution of the e2e suite
fc94f67
Serial execution has been the dominant wall-clock cost per matrix entry
since concurrency was disabled before conf (78048e0): ~128 tests at
~22 of 24 minutes on the Vercel lanes, and lately the slowest lane
cannot finish under its 30-minute job timeout on a slow runner day at
all. #2083 measured the concurrent suite at ~3x job wall-clock (4-5x on
the vitest phase) and identified what broke; its blockers are now fixed:
world-local writeExclusive is atomic (write-then-link), abort-fetch
tests are hermetic (#3618), the fibonacci tree fits the scheduler
(#3619), and source-map assertions are positive-only (#3620).
What this change adds is concurrency-safe per-test attribution. The
harness tracked runs and test names in module globals reset by a
beforeEach - under concurrency every test clobbered every other's
state, so a failing test dumped an unrelated sibling's diagnostics.
vitest's getCurrentTest() cannot substitute: it is a plain module
variable, wrong after any await. Instead an auto fixture - the one
place that receives the test's own context unambiguously - binds a
per-test state (name, tracked runs, the test's own skip) via
AsyncLocalStorage around each test body, and trackRun /
recordInfraEvent / requireFixture read it ambiently with no call-site
changes. The conformance gates skip through the bound state's skip, so
a mid-body requireFixture skips the right test. Sequential suites
(dev, agent, region) keep setupRunTracking's module-global fallback.
Full suite passes 137/137 concurrently against a local dev server in
under 2 minutes. A test that genuinely cannot share a deployment can
opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
[e2e] Bind per-test state via a task collector, not a fixture
f437a21
CI on the first concurrent run caught the fixture design failing in the
one place skip-correctness is load-bearing: vitest resolves fixtures in
a separate async context, so an AsyncLocalStorage store bound around
use() never reaches the test body. Every ambient read silently fell to
the module fallback - conformance-gated tests on the Python lane ran
instead of skipping and hard-failed on fixtures missing from the
manifest ('declared in e2e-conformance.json but not in the deployed
manifest').
The suite's test is now built with createTaskCollector, mirroring
vitest's own argument normalization, and the wrapper binds the state
around the handler call itself - a direct call stack, so propagation is
guaranteed rather than assumed. Validated by a collector-pattern ALS
experiment (per-test store after awaits, mid-body skip targets self,
each/skip/todo chains), a conformance-file simulation reproducing the
Python failure shape (fixture-missing tests now skip with the right
note), and a 137/137 concurrent local run.
Also raises RACE_WINNER_MAX_DURATION_MS 5s -> 8s: the winner takes 1s,
the loser 10s, and the bound only has to sit clearly below the loser -
concurrent-suite queue latency pushed observed winner durations to
~6.5s on loaded local-dev lanes (vite, first concurrent CI run), so 5s
flaked without catching anything 8s misses.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
alangenfeldforce-pushed
the
alangenfeld/e2e-concurrent-v2
branch
from
0556953 to
f437a21CompareAugust 19, 2026 15:05
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
📊 Workflow Benchmarks
commit c35ea2c · Wed, 19 Aug 2026 18:18:12 GMT · run logs
Backend: vercel · app: nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1380 (+381%) 🔻1517 🔴 (+30%) 🔻1557 🔴 (+16%) 🔻1624 🔴 (+3.9%)30
TTFSstream311 (+65%) 🔻1510 🔴 (+35%) 🔻1575 🔴 (+38%) 🔻1913 🔴 (+64%) 🔻30
TTFShook + stream1625 (+25%) 🔻1970 🔴 (+43%) 🔻2048 🔴 (+45%) 🔻2524 🔴 (+47%) 🔻30
Fan-out TTFSPromise.all(100 steps)1152 (+4.3%)2804 (+22%) 🔻3163 (+14%)3207 (+16%) 🔻10
Fan-out TTLSPromise.all(100 steps)5366 (+1.6%)8639 (+30%) 🔻10891 (+58%) 🔻19609 (+149%) 🔻10
STSO1020 steps (inline)117 (-11%)164 (-38%) 💚190 (-38%) 💚365 (-28%) 💚1019
WO1020 steps166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚1
CRTTfirst chunk (pooled)112 (+33%) 🔻186 (+46%) 🔻300 (-4.8%)436 (-30%) 💚28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)172 (+52%)176 (+7%)254 (+9%)457 (-2%)159 (-1%)10
size sweep (100/s, 160B-12KB)149 (+39%)165 (+7%)324 (+15%)518 (-24%)151 (-6%)10
replay gateway-gpt-5.4-nano-2000t (1x)126 (+13%)158 (-43%)240 (-67%)719 (-52%)353 (-41%)3
replay eve-gpt-5.6-sol-2000t (1x)135 (-14%)146 (+4%)186 (-3%)743 (+18%)523 (+7%)2
replay eve-gpt-5.6-sol-2000t (2x)138 (-73%)247 (-39%)355 (-87%)729 (-83%)569 (-6%)3
📈 STSO distribution vs main (inline / queue-hop histograms)
1020 steps (inline)
Cumulative STSO time: main 247209ms → this run 164994ms (Δ -82215ms, -33%)
100-150 ms █░░░░░░░░░░░░░░░░░░░░░░┃ main 6 this 474 +468
150-200 ms ██████████░░░░░░░░░░░░┃ main 205 this 464 +259
200-250 ms █┃██████████████████████ main 481 this 45 -436
250-300 ms ┃██████████ main 217 this 15 -202
300-350 ms ┃██ main 58 this 7 -51
350-400 ms ┃ main 20 this 8 -12
400-450 ms ┃ main 12 this 3 -9
450-500 ms ┃ main 9 this 1 -8
500-550 ms ┃ main 4 this 1 -3
550-600 ms ┃ main 2 this 0 -2
600-650 ms ┃ main 4 this 1 -3
850-900 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)
variant RTT 1ms→5s+ avg p50 p90 p99 n
control ······▂█▂···· 148.6 (+19%) 137 (+27%) 254 (+9%) 457 (-2%) 3000
sweep ······▃█▂▁··· 145.5 (+10%) 129 (+15%) 324 (+15%) 518 (-24%) 3000
gw 1x ·····▁▄█▁▁··· 132.2 (-25%) 116 (+5%) 240 (-67%) 719 (-52%) 5295
eve 1x ·····▁▅█▁▁··· 133.2 (+7%) 113 (+5%) 186 (-3%) 743 (+18%) 5186
eve 2x ·····▁▂█▃▁··· 173.2 (-55%) 139 (-19%) 355 (-87%) 729 (-83%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):
control █▄▃▂▄▂▄▁▁▂ 131–192ms
sweep ▇▃▂▁▃▃▄█▂▂ 124–181ms
gw 1x ▃█▄▂▂▄▁▂▃▄ 110–178ms
eve 1x ▂▃█▃▁▂▅▄▃█ 101–182ms
eve 2x █▁▂▂▁▄▅▅▄▅ 127–257ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):
sweep █▆▃▁▄▅█ 142–148ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):
control ▁▇▆▅▇▄█▄▆▄ 27–47ms
sweep ▁▆▆▆█▇▇▇▅▄ 31–61ms
gw 1x ▄▆▄▁▄█▃▄▃▅ 32–45ms
eve 1x ▄▆▆▇▁▄█▆▃█ 20–29ms
eve 2x █▁▃▂▃▅▃▁▄▅ 18–35ms

📜 Previous results (2)
c35ea2c
Wed, 19 Aug 2026 17:11:18 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1255 (+337%) 🔻1314 🔴 (+13%)1358 🔴 (+1.4%)1383 🔴 (-12%)30
TTFSstream241 (+28%) 🔻1334 🔴 (+20%) 🔻1399 🔴 (+23%) 🔻1681 🔴 (+44%) 🔻30
TTFShook + stream367 (-72%) 💚1748 🔴 (+27%) 🔻1778 🔴 (+26%) 🔻1952 🔴 (+14%)30
Fan-out TTFSPromise.all(100 steps)1392 (+26%) 🔻2936 (+28%) 🔻3132 (+13%)3151 (+14%)10
Fan-out TTLSPromise.all(100 steps)5963 (+13%)7084 (+6.4%)7237 (+5.1%)8436 (+7.0%)10
STSO1020 steps (inline)111 (-16%) 💚272 (+3.4%)320 (+5.3%)495 (-2.0%)1019
WO1020 steps245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)1
CRTTfirst chunk (pooled)98 (+17%) 🔻169 (+33%) 🔻398 (+26%) 🔻824 (+31%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)135 (+19%)434 (+163%)618 (+166%)1102 (+137%)617 (+287%)10
size sweep (100/s, 160B-12KB)149 (+39%)344 (+123%)540 (+92%)1044 (+54%)288 (+79%)10
replay gateway-gpt-5.4-nano-2000t (1x)127 (+13%)169 (-39%)290 (-60%)750 (-50%)524 (-13%)3
replay eve-gpt-5.6-sol-2000t (1x)206 (+32%)151 (+8%)246 (+28%)501 (-20%)380 (-22%)2
replay eve-gpt-5.6-sol-2000t (2x)153 (-70%)191 (-53%)315 (-89%)676 (-84%)444 (-27%)3
f437a21
Wed, 19 Aug 2026 15:26:31 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1341 (+367%) 🔻1586 🔴 (+36%) 🔻1690 🔴 (+26%) 🔻1885 🔴 (+21%) 🔻30
TTFSstream362 (+92%) 🔻1460 🔴 (+31%) 🔻1538 🔴 (+35%) 🔻1747 🔴 (+49%) 🔻30
TTFShook + stream1593 (+22%) 🔻1872 🔴 (+36%) 🔻1946 🔴 (+38%) 🔻2005 🔴 (+17%) 🔻30
Fan-out TTFSPromise.all(100 steps)1248 (+13%)2227 (-3.2%)2780 (+0.5%)3913 (+41%) 🔻10
Fan-out TTLSPromise.all(100 steps)5426 (+2.7%)11957 (+80%) 🔻22902 (+233%) 🔻24329 (+209%) 🔻10
STSO1020 steps (inline)111 (-16%) 💚164 (-38%) 💚187 (-38%) 💚285 (-44%) 💚1019
WO1020 steps159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚1
CRTTfirst chunk (pooled)96 (+14%)204 (+61%) 🔻341 (+8.3%)765 (+22%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)141 (+25%)184 (+12%)398 (+72%)735 (+58%)152 (-5%)10
size sweep (100/s, 160B-12KB)168 (+57%)155 (+1%)206 (-27%)408 (-40%)126 (-22%)10
replay gateway-gpt-5.4-nano-2000t (1x)316 (+182%)167 (-40%)225 (-69%)429 (-71%)293 (-51%)3
replay eve-gpt-5.6-sol-2000t (1x)161 (+3%)166 (+19%)217 (+13%)556 (-11%)270 (-45%)2
replay eve-gpt-5.6-sol-2000t (2x)162 (-68%)201 (-50%)264 (-91%)537 (-87%)177 (-71%)3
ℹ️ Metric definitions & methodology
Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.
The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.
The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.
Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.
Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)
Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles
Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041
🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600
All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).
Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
🧪 E2E Test Results
✅ All tests passed
⚠️ Flaky E2E Tests (passed on retry)
These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.
  • cancelRun via CLI - cancelling a running workflow (astro)
  • cancelRun via CLI - cancelling a running workflow (vite)
  • deploymentId: 'latest' is a no-op in non-Vercel worlds (nitro)
  • hookGetConflictThenStepParallelWorkflow - hook.getConflict() continuation step runs alongside other steps (nextjs-webpack)
🛠 Infra Events (absorbed by the harness)
Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.
  • cold-start-warmup · suite warmup (tanstack-start) · at 18:14:32Z · abandoned wrun_01M0DKNYTD6N9KWDRTEMB63XWM
  • run-pickup-stall · hookGetConflictWorkflow - hook.getConflict() resolves with the conflicting run when token is already registered (nuxt) · at 18:15:38Z · abandoned wrun_01M0DKRB1H8KCG69A7VFN9V5XN
E2E Test Summary
Summary
PassedFailedSkippedTotal
✅ ▲ Vercel Production347407384212
✅ 💻 Local Development381005584368
✅ 📦 Local Production381005584368
✅ 🐘 Local Postgres381005584368
✅ 🪟 Windows31200312
✅ 🌐 Cross-language Conformance90128137
✅ vercel-multi-region270027
Total152520254017792
Details by Category
✅ ▲ Vercel Production
AppPassedFailedSkipped
✅ astro-node128028
✅ astro-quickjs128028
✅ example-node128028
✅ example-quickjs128028
✅ express-node128028
✅ express-quickjs128028
✅ fastify-node128028
✅ fastify-quickjs128028
✅ hono-node128028
✅ hono-quickjs128028
✅ nest-node128028
✅ nest-quickjs128028
✅ nextjs-turbopack-node15303
✅ nextjs-turbopack-quickjs15303
✅ nextjs-webpack-node15303
✅ nextjs-webpack-quickjs15303
✅ nitro-node128028
✅ nitro-quickjs128028
✅ nuxt-node128028
✅ nuxt-quickjs128028
✅ python-node80148
✅ sveltekit-node14709
✅ sveltekit-quickjs14709
✅ tanstack-start-node128028
✅ tanstack-start-quickjs128028
✅ vite-node128028
✅ vite-quickjs128028
✅ 💻 Local Development
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 📦 Local Production
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🐘 Local Postgres
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🪟 Windows
AppPassedFailedSkipped
✅ nextjs-turbopack-node15600
✅ nextjs-turbopack-quickjs15600
✅ 🌐 Cross-language Conformance
AppPassedFailedSkipped
✅ python90128
✅ vercel-multi-region
AppPassedFailedSkipped
✅ nextjs-turbopack2700
📋 View full workflow run
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
Sim World
Simulated world deterministic testing for races. Traces
🟠 world-sim scenario book — 1 fail of 41 total
fence=per-spec
scenariooutcomeeventsvirtreplayviolations
✅ smoke-no-stepscompleted30msok0
✅ smoke-one-stepcompleted60msok0
✅ hook-at-step-startedcompleted120msok0
✅ hook-at-step-completedcompleted120msok0
✅ hook-at-hook-createdcompleted120msok0
✅ deadline-hook-winscompleted71.0hok0
✅ deadline-expirescompleted71.0hok0
✅ long-sleepcompleted1130.0dok0
✅ hook-never-arrivesstalled30msskipped0
✅ step-retries-twicecompleted102.0sok0
✅ parallel-stepscompleted90msok0
✅ hook-on-execution-statecompleted120msok0
✅ peek-hook-before-branchcompleted120msok0
✅ peek-hook-after-branchcompleted120msok0
✅ peek-hook-at-registrationcompleted120msok0
✅ race-hook-before-probecompleted120msok0
✅ race-hook-after-probecompleted120msok0
✅ race-duplicate-deliverycompleted130msok0
✅ attr-hook-before-stepcompleted110msok0
✅ attr-hook-after-stepcompleted110msok0
✅ attr-from-step-bodycompleted130msok0
✅ fork-hook-after-timeoutcompleted141.0mok0
✅ fork-hook-before-timeoutcompleted141.0mok0
✅ count-hook-after-timeoutcompleted171.0mok0
✅ count-hook-before-timeoutcompleted201.0mok0
✅ stale-read-step-count-forkcompleted201.0mok0
✅ stale-read-equal-step-countscompleted141.0mok0
✅ step-vs-step-forkcompleted120msok0
✅ step-vs-step-fork-fencedcompleted120msok0
✅ fence-catches-benign-directioncompleted125msok0
✅ in-flight-before-decisioncompleted171.0mok0
❌ in-flight-before-decision-countedcompleted171.0mok0
✅ in-flight-after-decisioncompleted192.0mok0
✅ stale-read-step-count-fork-fencedcompleted201.0mok0
✅ fork-hook-winscompleted131.0mok0
✅ fork-timeout-winscompleted131.0mok0
✅ unclaimed-payload-under-forkcompleted171.0mok0
✅ claimed-payload-under-forkcompleted171.0mok0
✅ writers-independent-step-bodiescompleted120msok0
✅ writers-scripted-tempocompleted120msok0
✅ cancel-mid-stepcancelled70msskipped0
Full trace: world-sim.txt
@VaguelySeriousVaguelySerious
mentioned this pull request
Aug 19, 2026
 Merged

@alangenfeld
[e2e] Delegate to the suite collector so describe.concurrent applies
ef36a7e
The previous commit's collector called getCurrentSuite().task() directly,
which skips the suite collector's Object.assign({}, suiteOptions,
options) - the merge that carries describe.concurrent onto each test. The
suite therefore still ran sequentially, and CI proved it: every Vercel
Prod lane matched the serial baseline minute-for-minute
(nextjs-turbopack quickjs 25m vs 25m), so a fully green run said nothing
about concurrency. The same slowdown showed up locally as a full-suite
run taking 6m15s where the earlier concurrent one took 1m58s, which I
had misread as server slowness.
The collector now wraps the handler and hands the call to
getCurrentSuite().test.fn - exactly what vitest's own top-level test
does - so suite-option inheritance is vitest's code path again, while
the wrapper still owns the handler call stack that makes the
AsyncLocalStorage binding reach the test body.
Verified with an experiment asserting both properties at once: four
300ms tests plus a 600ms observer finish in 602ms (1800ms if
sequential), max-in-flight > 1, and each test still reads its own store
after awaits.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
vercel[bot]
vercelBot
reviewed
Aug 19, 2026
View reviewed changes
Comment threadpackages/core/e2e/e2e.test.ts
@alangenfeld
[e2e] Actually call the load-observability hooks
c35ea2c
The previous commit added the counters and the summary but left the call
sites out of the handler wrapper and afterAll, so nothing was collected
and no summary printed. Caught by review.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@github-actions
github-actionsBot
commented
Aug 20, 2026

Copy link


Copy Markdown

Contributor
No backport to stable for 8c29b85 (AI decision).
This is test-harness enhancement work, not a fix: it re-enables describe.concurrent for the e2e suite to cut CI wall clock, adds a new load-summary reporting facility, a new WORKFLOW_E2E_MAX_CONCURRENCY config knob, and reworks per-test state onto AsyncLocalStorage. Nothing here repairs broken or flaky behavior on stable — it trades a known-serial, known-green suite for a concurrency model whose safe value was tuned against main's lanes, which is exactly the kind of risk a maintenance branch is avoiding. The RACE_WINNER_MAX_DURATION_MS 5s→8s bump is the only flake-adjacent piece, and it is only needed because of the concurrency this commit introduces.
To override, re-run the Backport to stable workflow manually via workflow_dispatch and paste this commit SHA into the ref input:
8c29b85322dc2279b6274391cd14a1483eb969c3

Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment

Reviewers

@vercelvercel[bot]vercel[bot] left review comments
@TooTallNateTooTallNateTooTallNate approved these changes

Assignees


No one assigned


Labels


None yet


Projects


None yet


Milestone


No milestone


Development

Successfully merging this pull request may close these issues.

3 participants

@alangenfeld@TooTallNate@VaguelySerious

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Auto-enable theater mode on YouTube
(function() {
function tryTheater() {
var btn = document.querySelector('button[aria-label="Theater mode"], ytd-player #player button[title="Theater mode"]');
if (btn && !btn.classList.contains('activated')) {
btn.click();
}
}
// Try immediately
tryTheater();
// Try after navigation (SPA)
var lastUrl = location.href;
setInterval(function() {
if (location.href !== lastUrl) {
lastUrl = location.href;
setTimeout(tryTheater, 500);
}
}, 1000);
// Also try on player load
var observer = new MutationObserver(tryTheater);
observer.observe(document.body, { childList: true, subtree: true });
})();
}
} catch(__e) { console.warn('[Userscript:YouTube Theater Mode Default]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '
[e2e] Re-enable concurrent execution of the e2e suite by alangenfeld · Pull Request #3658 · vercel/workflow · GitHub
Skip to content
[e2e] Re-enable concurrent execution of the e2e suite - #3658
Merged
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Aug 20, 2026
Merged
[e2e] Re-enable concurrent execution of the e2e suite#3658
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Conversation
@alangenfeld
@alangenfeldalangenfeld
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Collaborator
Summary & Motivation
Serial execution was the dominant wall-clock cost per matrix entry (~22 of 24 minutes on the Vercel lanes), and the blockers that forced it in 78048e0 are fixed.
  • Per-test harness state (name, tracked runs, skip) binds through AsyncLocalStorage instead of module globals, so a failing test dumps its own diagnostics and a mid-body conformance gate skips the right test. The suite's test is a createTaskCollector wrapper that delegates to the enclosing suite's collector, which is both how the binding gets a direct call stack to the handler and how describe.concurrent inheritance reaches each test.
  • Each lane ends with a load summary — test and CLI child counts, peak concurrency of each, median test wall time, CLI share of summed test time, ten slowest tests — since under concurrency the question is where per-test latency went, not pass/fail.
  • maxConcurrency is tunable via WORKFLOW_E2E_MAX_CONCURRENCY, default unchanged at 5. Higher values starve CLI children on lanes that also host the app and Postgres.
  • RACE_WINNER_MAX_DURATION_MS goes 5s → 8s; queue latency under concurrency pushed observed winner durations to ~6.5s, still far below the 10s loser.
A test that genuinely cannot share a deployment can opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Test Plan
Added a unit test for interleaved per-test attribution; five consecutive full-matrix CI runs at maxConcurrency=5 were green (162/162) with Vercel-prod lanes at 5-7 min against 22-25 min serial and median test time unchanged, and a maxConcurrency=8 probe established 5 as the value safe on every lane type.
@changeset-bot
changeset-botBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

⚠️ No Changeset found
Latest commit: c35ea2c
Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.
This PR includes no changesets
When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types
Click here to learn what changesets are, and how to add one.
Click here if you're a maintainer who wants to add a changeset to this PR
@vercel
vercelBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
The latest updates on your projects. Learn more about Vercel for GitHub.
ProjectDeploymentActionsUpdated (UTC)
example-nextjs-workflow-turbopackReadyReadyPreviewAug 19, 2026 5:45pm
example-nextjs-workflow-webpackReadyReadyPreviewAug 19, 2026 5:45pm
example-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-astro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-express-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-fastify-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-hono-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nestjs-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nitro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nuxt-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-python-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-sveltekit-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-tanstack-start-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-vite-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workflow-docsReadyReadyPreview, v0Aug 19, 2026 5:45pm
workflow-swc-playgroundReadyReadyPreviewAug 19, 2026 5:45pm
workflow-tarballsReadyReadyPreviewAug 19, 2026 5:45pm
workflow-webReadyReadyPreviewAug 19, 2026 5:45pm
@alangenfeld
[e2e] Re-enable concurrent execution of the e2e suite
fc94f67
Serial execution has been the dominant wall-clock cost per matrix entry
since concurrency was disabled before conf (78048e0): ~128 tests at
~22 of 24 minutes on the Vercel lanes, and lately the slowest lane
cannot finish under its 30-minute job timeout on a slow runner day at
all. #2083 measured the concurrent suite at ~3x job wall-clock (4-5x on
the vitest phase) and identified what broke; its blockers are now fixed:
world-local writeExclusive is atomic (write-then-link), abort-fetch
tests are hermetic (#3618), the fibonacci tree fits the scheduler
(#3619), and source-map assertions are positive-only (#3620).
What this change adds is concurrency-safe per-test attribution. The
harness tracked runs and test names in module globals reset by a
beforeEach - under concurrency every test clobbered every other's
state, so a failing test dumped an unrelated sibling's diagnostics.
vitest's getCurrentTest() cannot substitute: it is a plain module
variable, wrong after any await. Instead an auto fixture - the one
place that receives the test's own context unambiguously - binds a
per-test state (name, tracked runs, the test's own skip) via
AsyncLocalStorage around each test body, and trackRun /
recordInfraEvent / requireFixture read it ambiently with no call-site
changes. The conformance gates skip through the bound state's skip, so
a mid-body requireFixture skips the right test. Sequential suites
(dev, agent, region) keep setupRunTracking's module-global fallback.
Full suite passes 137/137 concurrently against a local dev server in
under 2 minutes. A test that genuinely cannot share a deployment can
opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
[e2e] Bind per-test state via a task collector, not a fixture
f437a21
CI on the first concurrent run caught the fixture design failing in the
one place skip-correctness is load-bearing: vitest resolves fixtures in
a separate async context, so an AsyncLocalStorage store bound around
use() never reaches the test body. Every ambient read silently fell to
the module fallback - conformance-gated tests on the Python lane ran
instead of skipping and hard-failed on fixtures missing from the
manifest ('declared in e2e-conformance.json but not in the deployed
manifest').
The suite's test is now built with createTaskCollector, mirroring
vitest's own argument normalization, and the wrapper binds the state
around the handler call itself - a direct call stack, so propagation is
guaranteed rather than assumed. Validated by a collector-pattern ALS
experiment (per-test store after awaits, mid-body skip targets self,
each/skip/todo chains), a conformance-file simulation reproducing the
Python failure shape (fixture-missing tests now skip with the right
note), and a 137/137 concurrent local run.
Also raises RACE_WINNER_MAX_DURATION_MS 5s -> 8s: the winner takes 1s,
the loser 10s, and the bound only has to sit clearly below the loser -
concurrent-suite queue latency pushed observed winner durations to
~6.5s on loaded local-dev lanes (vite, first concurrent CI run), so 5s
flaked without catching anything 8s misses.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
alangenfeldforce-pushed
the
alangenfeld/e2e-concurrent-v2
branch
from
0556953 to
f437a21CompareAugust 19, 2026 15:05
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
📊 Workflow Benchmarks
commit c35ea2c · Wed, 19 Aug 2026 18:18:12 GMT · run logs
Backend: vercel · app: nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1380 (+381%) 🔻1517 🔴 (+30%) 🔻1557 🔴 (+16%) 🔻1624 🔴 (+3.9%)30
TTFSstream311 (+65%) 🔻1510 🔴 (+35%) 🔻1575 🔴 (+38%) 🔻1913 🔴 (+64%) 🔻30
TTFShook + stream1625 (+25%) 🔻1970 🔴 (+43%) 🔻2048 🔴 (+45%) 🔻2524 🔴 (+47%) 🔻30
Fan-out TTFSPromise.all(100 steps)1152 (+4.3%)2804 (+22%) 🔻3163 (+14%)3207 (+16%) 🔻10
Fan-out TTLSPromise.all(100 steps)5366 (+1.6%)8639 (+30%) 🔻10891 (+58%) 🔻19609 (+149%) 🔻10
STSO1020 steps (inline)117 (-11%)164 (-38%) 💚190 (-38%) 💚365 (-28%) 💚1019
WO1020 steps166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚1
CRTTfirst chunk (pooled)112 (+33%) 🔻186 (+46%) 🔻300 (-4.8%)436 (-30%) 💚28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)172 (+52%)176 (+7%)254 (+9%)457 (-2%)159 (-1%)10
size sweep (100/s, 160B-12KB)149 (+39%)165 (+7%)324 (+15%)518 (-24%)151 (-6%)10
replay gateway-gpt-5.4-nano-2000t (1x)126 (+13%)158 (-43%)240 (-67%)719 (-52%)353 (-41%)3
replay eve-gpt-5.6-sol-2000t (1x)135 (-14%)146 (+4%)186 (-3%)743 (+18%)523 (+7%)2
replay eve-gpt-5.6-sol-2000t (2x)138 (-73%)247 (-39%)355 (-87%)729 (-83%)569 (-6%)3
📈 STSO distribution vs main (inline / queue-hop histograms)
1020 steps (inline)
Cumulative STSO time: main 247209ms → this run 164994ms (Δ -82215ms, -33%)
100-150 ms █░░░░░░░░░░░░░░░░░░░░░░┃ main 6 this 474 +468
150-200 ms ██████████░░░░░░░░░░░░┃ main 205 this 464 +259
200-250 ms █┃██████████████████████ main 481 this 45 -436
250-300 ms ┃██████████ main 217 this 15 -202
300-350 ms ┃██ main 58 this 7 -51
350-400 ms ┃ main 20 this 8 -12
400-450 ms ┃ main 12 this 3 -9
450-500 ms ┃ main 9 this 1 -8
500-550 ms ┃ main 4 this 1 -3
550-600 ms ┃ main 2 this 0 -2
600-650 ms ┃ main 4 this 1 -3
850-900 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)
variant RTT 1ms→5s+ avg p50 p90 p99 n
control ······▂█▂···· 148.6 (+19%) 137 (+27%) 254 (+9%) 457 (-2%) 3000
sweep ······▃█▂▁··· 145.5 (+10%) 129 (+15%) 324 (+15%) 518 (-24%) 3000
gw 1x ·····▁▄█▁▁··· 132.2 (-25%) 116 (+5%) 240 (-67%) 719 (-52%) 5295
eve 1x ·····▁▅█▁▁··· 133.2 (+7%) 113 (+5%) 186 (-3%) 743 (+18%) 5186
eve 2x ·····▁▂█▃▁··· 173.2 (-55%) 139 (-19%) 355 (-87%) 729 (-83%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):
control █▄▃▂▄▂▄▁▁▂ 131–192ms
sweep ▇▃▂▁▃▃▄█▂▂ 124–181ms
gw 1x ▃█▄▂▂▄▁▂▃▄ 110–178ms
eve 1x ▂▃█▃▁▂▅▄▃█ 101–182ms
eve 2x █▁▂▂▁▄▅▅▄▅ 127–257ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):
sweep █▆▃▁▄▅█ 142–148ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):
control ▁▇▆▅▇▄█▄▆▄ 27–47ms
sweep ▁▆▆▆█▇▇▇▅▄ 31–61ms
gw 1x ▄▆▄▁▄█▃▄▃▅ 32–45ms
eve 1x ▄▆▆▇▁▄█▆▃█ 20–29ms
eve 2x █▁▃▂▃▅▃▁▄▅ 18–35ms

📜 Previous results (2)
c35ea2c
Wed, 19 Aug 2026 17:11:18 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1255 (+337%) 🔻1314 🔴 (+13%)1358 🔴 (+1.4%)1383 🔴 (-12%)30
TTFSstream241 (+28%) 🔻1334 🔴 (+20%) 🔻1399 🔴 (+23%) 🔻1681 🔴 (+44%) 🔻30
TTFShook + stream367 (-72%) 💚1748 🔴 (+27%) 🔻1778 🔴 (+26%) 🔻1952 🔴 (+14%)30
Fan-out TTFSPromise.all(100 steps)1392 (+26%) 🔻2936 (+28%) 🔻3132 (+13%)3151 (+14%)10
Fan-out TTLSPromise.all(100 steps)5963 (+13%)7084 (+6.4%)7237 (+5.1%)8436 (+7.0%)10
STSO1020 steps (inline)111 (-16%) 💚272 (+3.4%)320 (+5.3%)495 (-2.0%)1019
WO1020 steps245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)1
CRTTfirst chunk (pooled)98 (+17%) 🔻169 (+33%) 🔻398 (+26%) 🔻824 (+31%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)135 (+19%)434 (+163%)618 (+166%)1102 (+137%)617 (+287%)10
size sweep (100/s, 160B-12KB)149 (+39%)344 (+123%)540 (+92%)1044 (+54%)288 (+79%)10
replay gateway-gpt-5.4-nano-2000t (1x)127 (+13%)169 (-39%)290 (-60%)750 (-50%)524 (-13%)3
replay eve-gpt-5.6-sol-2000t (1x)206 (+32%)151 (+8%)246 (+28%)501 (-20%)380 (-22%)2
replay eve-gpt-5.6-sol-2000t (2x)153 (-70%)191 (-53%)315 (-89%)676 (-84%)444 (-27%)3
f437a21
Wed, 19 Aug 2026 15:26:31 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1341 (+367%) 🔻1586 🔴 (+36%) 🔻1690 🔴 (+26%) 🔻1885 🔴 (+21%) 🔻30
TTFSstream362 (+92%) 🔻1460 🔴 (+31%) 🔻1538 🔴 (+35%) 🔻1747 🔴 (+49%) 🔻30
TTFShook + stream1593 (+22%) 🔻1872 🔴 (+36%) 🔻1946 🔴 (+38%) 🔻2005 🔴 (+17%) 🔻30
Fan-out TTFSPromise.all(100 steps)1248 (+13%)2227 (-3.2%)2780 (+0.5%)3913 (+41%) 🔻10
Fan-out TTLSPromise.all(100 steps)5426 (+2.7%)11957 (+80%) 🔻22902 (+233%) 🔻24329 (+209%) 🔻10
STSO1020 steps (inline)111 (-16%) 💚164 (-38%) 💚187 (-38%) 💚285 (-44%) 💚1019
WO1020 steps159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚1
CRTTfirst chunk (pooled)96 (+14%)204 (+61%) 🔻341 (+8.3%)765 (+22%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)141 (+25%)184 (+12%)398 (+72%)735 (+58%)152 (-5%)10
size sweep (100/s, 160B-12KB)168 (+57%)155 (+1%)206 (-27%)408 (-40%)126 (-22%)10
replay gateway-gpt-5.4-nano-2000t (1x)316 (+182%)167 (-40%)225 (-69%)429 (-71%)293 (-51%)3
replay eve-gpt-5.6-sol-2000t (1x)161 (+3%)166 (+19%)217 (+13%)556 (-11%)270 (-45%)2
replay eve-gpt-5.6-sol-2000t (2x)162 (-68%)201 (-50%)264 (-91%)537 (-87%)177 (-71%)3
ℹ️ Metric definitions & methodology
Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.
The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.
The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.
Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.
Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)
Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles
Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041
🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600
All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).
Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
🧪 E2E Test Results
✅ All tests passed
⚠️ Flaky E2E Tests (passed on retry)
These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.
  • cancelRun via CLI - cancelling a running workflow (astro)
  • cancelRun via CLI - cancelling a running workflow (vite)
  • deploymentId: 'latest' is a no-op in non-Vercel worlds (nitro)
  • hookGetConflictThenStepParallelWorkflow - hook.getConflict() continuation step runs alongside other steps (nextjs-webpack)
🛠 Infra Events (absorbed by the harness)
Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.
  • cold-start-warmup · suite warmup (tanstack-start) · at 18:14:32Z · abandoned wrun_01M0DKNYTD6N9KWDRTEMB63XWM
  • run-pickup-stall · hookGetConflictWorkflow - hook.getConflict() resolves with the conflicting run when token is already registered (nuxt) · at 18:15:38Z · abandoned wrun_01M0DKRB1H8KCG69A7VFN9V5XN
E2E Test Summary
Summary
PassedFailedSkippedTotal
✅ ▲ Vercel Production347407384212
✅ 💻 Local Development381005584368
✅ 📦 Local Production381005584368
✅ 🐘 Local Postgres381005584368
✅ 🪟 Windows31200312
✅ 🌐 Cross-language Conformance90128137
✅ vercel-multi-region270027
Total152520254017792
Details by Category
✅ ▲ Vercel Production
AppPassedFailedSkipped
✅ astro-node128028
✅ astro-quickjs128028
✅ example-node128028
✅ example-quickjs128028
✅ express-node128028
✅ express-quickjs128028
✅ fastify-node128028
✅ fastify-quickjs128028
✅ hono-node128028
✅ hono-quickjs128028
✅ nest-node128028
✅ nest-quickjs128028
✅ nextjs-turbopack-node15303
✅ nextjs-turbopack-quickjs15303
✅ nextjs-webpack-node15303
✅ nextjs-webpack-quickjs15303
✅ nitro-node128028
✅ nitro-quickjs128028
✅ nuxt-node128028
✅ nuxt-quickjs128028
✅ python-node80148
✅ sveltekit-node14709
✅ sveltekit-quickjs14709
✅ tanstack-start-node128028
✅ tanstack-start-quickjs128028
✅ vite-node128028
✅ vite-quickjs128028
✅ 💻 Local Development
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 📦 Local Production
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🐘 Local Postgres
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🪟 Windows
AppPassedFailedSkipped
✅ nextjs-turbopack-node15600
✅ nextjs-turbopack-quickjs15600
✅ 🌐 Cross-language Conformance
AppPassedFailedSkipped
✅ python90128
✅ vercel-multi-region
AppPassedFailedSkipped
✅ nextjs-turbopack2700
📋 View full workflow run
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
Sim World
Simulated world deterministic testing for races. Traces
🟠 world-sim scenario book — 1 fail of 41 total
fence=per-spec
scenariooutcomeeventsvirtreplayviolations
✅ smoke-no-stepscompleted30msok0
✅ smoke-one-stepcompleted60msok0
✅ hook-at-step-startedcompleted120msok0
✅ hook-at-step-completedcompleted120msok0
✅ hook-at-hook-createdcompleted120msok0
✅ deadline-hook-winscompleted71.0hok0
✅ deadline-expirescompleted71.0hok0
✅ long-sleepcompleted1130.0dok0
✅ hook-never-arrivesstalled30msskipped0
✅ step-retries-twicecompleted102.0sok0
✅ parallel-stepscompleted90msok0
✅ hook-on-execution-statecompleted120msok0
✅ peek-hook-before-branchcompleted120msok0
✅ peek-hook-after-branchcompleted120msok0
✅ peek-hook-at-registrationcompleted120msok0
✅ race-hook-before-probecompleted120msok0
✅ race-hook-after-probecompleted120msok0
✅ race-duplicate-deliverycompleted130msok0
✅ attr-hook-before-stepcompleted110msok0
✅ attr-hook-after-stepcompleted110msok0
✅ attr-from-step-bodycompleted130msok0
✅ fork-hook-after-timeoutcompleted141.0mok0
✅ fork-hook-before-timeoutcompleted141.0mok0
✅ count-hook-after-timeoutcompleted171.0mok0
✅ count-hook-before-timeoutcompleted201.0mok0
✅ stale-read-step-count-forkcompleted201.0mok0
✅ stale-read-equal-step-countscompleted141.0mok0
✅ step-vs-step-forkcompleted120msok0
✅ step-vs-step-fork-fencedcompleted120msok0
✅ fence-catches-benign-directioncompleted125msok0
✅ in-flight-before-decisioncompleted171.0mok0
❌ in-flight-before-decision-countedcompleted171.0mok0
✅ in-flight-after-decisioncompleted192.0mok0
✅ stale-read-step-count-fork-fencedcompleted201.0mok0
✅ fork-hook-winscompleted131.0mok0
✅ fork-timeout-winscompleted131.0mok0
✅ unclaimed-payload-under-forkcompleted171.0mok0
✅ claimed-payload-under-forkcompleted171.0mok0
✅ writers-independent-step-bodiescompleted120msok0
✅ writers-scripted-tempocompleted120msok0
✅ cancel-mid-stepcancelled70msskipped0
Full trace: world-sim.txt
@VaguelySeriousVaguelySerious
mentioned this pull request
Aug 19, 2026
 Merged

@alangenfeld
[e2e] Delegate to the suite collector so describe.concurrent applies
ef36a7e
The previous commit's collector called getCurrentSuite().task() directly,
which skips the suite collector's Object.assign({}, suiteOptions,
options) - the merge that carries describe.concurrent onto each test. The
suite therefore still ran sequentially, and CI proved it: every Vercel
Prod lane matched the serial baseline minute-for-minute
(nextjs-turbopack quickjs 25m vs 25m), so a fully green run said nothing
about concurrency. The same slowdown showed up locally as a full-suite
run taking 6m15s where the earlier concurrent one took 1m58s, which I
had misread as server slowness.
The collector now wraps the handler and hands the call to
getCurrentSuite().test.fn - exactly what vitest's own top-level test
does - so suite-option inheritance is vitest's code path again, while
the wrapper still owns the handler call stack that makes the
AsyncLocalStorage binding reach the test body.
Verified with an experiment asserting both properties at once: four
300ms tests plus a 600ms observer finish in 602ms (1800ms if
sequential), max-in-flight > 1, and each test still reads its own store
after awaits.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
vercel[bot]
vercelBot
reviewed
Aug 19, 2026
View reviewed changes
Comment threadpackages/core/e2e/e2e.test.ts
@alangenfeld
[e2e] Actually call the load-observability hooks
c35ea2c
The previous commit added the counters and the summary but left the call
sites out of the handler wrapper and afterAll, so nothing was collected
and no summary printed. Caught by review.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@github-actions
github-actionsBot
commented
Aug 20, 2026

Copy link


Copy Markdown

Contributor
No backport to stable for 8c29b85 (AI decision).
This is test-harness enhancement work, not a fix: it re-enables describe.concurrent for the e2e suite to cut CI wall clock, adds a new load-summary reporting facility, a new WORKFLOW_E2E_MAX_CONCURRENCY config knob, and reworks per-test state onto AsyncLocalStorage. Nothing here repairs broken or flaky behavior on stable — it trades a known-serial, known-green suite for a concurrency model whose safe value was tuned against main's lanes, which is exactly the kind of risk a maintenance branch is avoiding. The RACE_WINNER_MAX_DURATION_MS 5s→8s bump is the only flake-adjacent piece, and it is only needed because of the concurrency this commit introduces.
To override, re-run the Backport to stable workflow manually via workflow_dispatch and paste this commit SHA into the ref input:
8c29b85322dc2279b6274391cd14a1483eb969c3

Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment

Reviewers

@vercelvercel[bot]vercel[bot] left review comments
@TooTallNateTooTallNateTooTallNate approved these changes

Assignees


No one assigned


Labels


None yet


Projects


None yet


Milestone


No milestone


Development

Successfully merging this pull request may close these issues.

3 participants

@alangenfeld@TooTallNate@VaguelySerious

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Remove or un-stick sticky/fixed headers that block content
(function() {
function unstick() {
document.querySelectorAll('header, nav, [role="banner"], .header, .navbar, .sticky, .fixed-top, [style*="position: fixed"], [style*="position:sticky"]').forEach(function(el) {
if (el.style.position === 'fixed' || el.style.position === 'sticky' || getComputedStyle(el).position === 'fixed' || getComputedStyle(el).position === 'sticky') {
el.style.position = 'static';
el.style.top = 'auto';
el.style.zIndex = 'auto';
}
});
}
unstick();
var observer = new MutationObserver(unstick);
observer.observe(document.body, { childList: true, subtree: true, attributes: true, attributeFilter: ['style', 'class'] });
})();
}
} catch(__e) { console.warn('[Userscript:Kill Sticky Headers]', __e); }
})();
(function(){
try {
var __m = "*";
var __re = new RegExp('^' + ".*" + '
[e2e] Re-enable concurrent execution of the e2e suite by alangenfeld · Pull Request #3658 · vercel/workflow · GitHub
Skip to content
[e2e] Re-enable concurrent execution of the e2e suite - #3658
Merged
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Aug 20, 2026
Merged
[e2e] Re-enable concurrent execution of the e2e suite#3658
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Conversation
@alangenfeld
@alangenfeldalangenfeld
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Collaborator
Summary & Motivation
Serial execution was the dominant wall-clock cost per matrix entry (~22 of 24 minutes on the Vercel lanes), and the blockers that forced it in 78048e0 are fixed.
  • Per-test harness state (name, tracked runs, skip) binds through AsyncLocalStorage instead of module globals, so a failing test dumps its own diagnostics and a mid-body conformance gate skips the right test. The suite's test is a createTaskCollector wrapper that delegates to the enclosing suite's collector, which is both how the binding gets a direct call stack to the handler and how describe.concurrent inheritance reaches each test.
  • Each lane ends with a load summary — test and CLI child counts, peak concurrency of each, median test wall time, CLI share of summed test time, ten slowest tests — since under concurrency the question is where per-test latency went, not pass/fail.
  • maxConcurrency is tunable via WORKFLOW_E2E_MAX_CONCURRENCY, default unchanged at 5. Higher values starve CLI children on lanes that also host the app and Postgres.
  • RACE_WINNER_MAX_DURATION_MS goes 5s → 8s; queue latency under concurrency pushed observed winner durations to ~6.5s, still far below the 10s loser.
A test that genuinely cannot share a deployment can opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Test Plan
Added a unit test for interleaved per-test attribution; five consecutive full-matrix CI runs at maxConcurrency=5 were green (162/162) with Vercel-prod lanes at 5-7 min against 22-25 min serial and median test time unchanged, and a maxConcurrency=8 probe established 5 as the value safe on every lane type.
@changeset-bot
changeset-botBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

⚠️ No Changeset found
Latest commit: c35ea2c
Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.
This PR includes no changesets
When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types
Click here to learn what changesets are, and how to add one.
Click here if you're a maintainer who wants to add a changeset to this PR
@vercel
vercelBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
The latest updates on your projects. Learn more about Vercel for GitHub.
ProjectDeploymentActionsUpdated (UTC)
example-nextjs-workflow-turbopackReadyReadyPreviewAug 19, 2026 5:45pm
example-nextjs-workflow-webpackReadyReadyPreviewAug 19, 2026 5:45pm
example-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-astro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-express-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-fastify-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-hono-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nestjs-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nitro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nuxt-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-python-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-sveltekit-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-tanstack-start-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-vite-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workflow-docsReadyReadyPreview, v0Aug 19, 2026 5:45pm
workflow-swc-playgroundReadyReadyPreviewAug 19, 2026 5:45pm
workflow-tarballsReadyReadyPreviewAug 19, 2026 5:45pm
workflow-webReadyReadyPreviewAug 19, 2026 5:45pm
@alangenfeld
[e2e] Re-enable concurrent execution of the e2e suite
fc94f67
Serial execution has been the dominant wall-clock cost per matrix entry
since concurrency was disabled before conf (78048e0): ~128 tests at
~22 of 24 minutes on the Vercel lanes, and lately the slowest lane
cannot finish under its 30-minute job timeout on a slow runner day at
all. #2083 measured the concurrent suite at ~3x job wall-clock (4-5x on
the vitest phase) and identified what broke; its blockers are now fixed:
world-local writeExclusive is atomic (write-then-link), abort-fetch
tests are hermetic (#3618), the fibonacci tree fits the scheduler
(#3619), and source-map assertions are positive-only (#3620).
What this change adds is concurrency-safe per-test attribution. The
harness tracked runs and test names in module globals reset by a
beforeEach - under concurrency every test clobbered every other's
state, so a failing test dumped an unrelated sibling's diagnostics.
vitest's getCurrentTest() cannot substitute: it is a plain module
variable, wrong after any await. Instead an auto fixture - the one
place that receives the test's own context unambiguously - binds a
per-test state (name, tracked runs, the test's own skip) via
AsyncLocalStorage around each test body, and trackRun /
recordInfraEvent / requireFixture read it ambiently with no call-site
changes. The conformance gates skip through the bound state's skip, so
a mid-body requireFixture skips the right test. Sequential suites
(dev, agent, region) keep setupRunTracking's module-global fallback.
Full suite passes 137/137 concurrently against a local dev server in
under 2 minutes. A test that genuinely cannot share a deployment can
opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
[e2e] Bind per-test state via a task collector, not a fixture
f437a21
CI on the first concurrent run caught the fixture design failing in the
one place skip-correctness is load-bearing: vitest resolves fixtures in
a separate async context, so an AsyncLocalStorage store bound around
use() never reaches the test body. Every ambient read silently fell to
the module fallback - conformance-gated tests on the Python lane ran
instead of skipping and hard-failed on fixtures missing from the
manifest ('declared in e2e-conformance.json but not in the deployed
manifest').
The suite's test is now built with createTaskCollector, mirroring
vitest's own argument normalization, and the wrapper binds the state
around the handler call itself - a direct call stack, so propagation is
guaranteed rather than assumed. Validated by a collector-pattern ALS
experiment (per-test store after awaits, mid-body skip targets self,
each/skip/todo chains), a conformance-file simulation reproducing the
Python failure shape (fixture-missing tests now skip with the right
note), and a 137/137 concurrent local run.
Also raises RACE_WINNER_MAX_DURATION_MS 5s -> 8s: the winner takes 1s,
the loser 10s, and the bound only has to sit clearly below the loser -
concurrent-suite queue latency pushed observed winner durations to
~6.5s on loaded local-dev lanes (vite, first concurrent CI run), so 5s
flaked without catching anything 8s misses.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
alangenfeldforce-pushed
the
alangenfeld/e2e-concurrent-v2
branch
from
0556953 to
f437a21CompareAugust 19, 2026 15:05
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
📊 Workflow Benchmarks
commit c35ea2c · Wed, 19 Aug 2026 18:18:12 GMT · run logs
Backend: vercel · app: nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1380 (+381%) 🔻1517 🔴 (+30%) 🔻1557 🔴 (+16%) 🔻1624 🔴 (+3.9%)30
TTFSstream311 (+65%) 🔻1510 🔴 (+35%) 🔻1575 🔴 (+38%) 🔻1913 🔴 (+64%) 🔻30
TTFShook + stream1625 (+25%) 🔻1970 🔴 (+43%) 🔻2048 🔴 (+45%) 🔻2524 🔴 (+47%) 🔻30
Fan-out TTFSPromise.all(100 steps)1152 (+4.3%)2804 (+22%) 🔻3163 (+14%)3207 (+16%) 🔻10
Fan-out TTLSPromise.all(100 steps)5366 (+1.6%)8639 (+30%) 🔻10891 (+58%) 🔻19609 (+149%) 🔻10
STSO1020 steps (inline)117 (-11%)164 (-38%) 💚190 (-38%) 💚365 (-28%) 💚1019
WO1020 steps166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚1
CRTTfirst chunk (pooled)112 (+33%) 🔻186 (+46%) 🔻300 (-4.8%)436 (-30%) 💚28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)172 (+52%)176 (+7%)254 (+9%)457 (-2%)159 (-1%)10
size sweep (100/s, 160B-12KB)149 (+39%)165 (+7%)324 (+15%)518 (-24%)151 (-6%)10
replay gateway-gpt-5.4-nano-2000t (1x)126 (+13%)158 (-43%)240 (-67%)719 (-52%)353 (-41%)3
replay eve-gpt-5.6-sol-2000t (1x)135 (-14%)146 (+4%)186 (-3%)743 (+18%)523 (+7%)2
replay eve-gpt-5.6-sol-2000t (2x)138 (-73%)247 (-39%)355 (-87%)729 (-83%)569 (-6%)3
📈 STSO distribution vs main (inline / queue-hop histograms)
1020 steps (inline)
Cumulative STSO time: main 247209ms → this run 164994ms (Δ -82215ms, -33%)
100-150 ms █░░░░░░░░░░░░░░░░░░░░░░┃ main 6 this 474 +468
150-200 ms ██████████░░░░░░░░░░░░┃ main 205 this 464 +259
200-250 ms █┃██████████████████████ main 481 this 45 -436
250-300 ms ┃██████████ main 217 this 15 -202
300-350 ms ┃██ main 58 this 7 -51
350-400 ms ┃ main 20 this 8 -12
400-450 ms ┃ main 12 this 3 -9
450-500 ms ┃ main 9 this 1 -8
500-550 ms ┃ main 4 this 1 -3
550-600 ms ┃ main 2 this 0 -2
600-650 ms ┃ main 4 this 1 -3
850-900 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)
variant RTT 1ms→5s+ avg p50 p90 p99 n
control ······▂█▂···· 148.6 (+19%) 137 (+27%) 254 (+9%) 457 (-2%) 3000
sweep ······▃█▂▁··· 145.5 (+10%) 129 (+15%) 324 (+15%) 518 (-24%) 3000
gw 1x ·····▁▄█▁▁··· 132.2 (-25%) 116 (+5%) 240 (-67%) 719 (-52%) 5295
eve 1x ·····▁▅█▁▁··· 133.2 (+7%) 113 (+5%) 186 (-3%) 743 (+18%) 5186
eve 2x ·····▁▂█▃▁··· 173.2 (-55%) 139 (-19%) 355 (-87%) 729 (-83%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):
control █▄▃▂▄▂▄▁▁▂ 131–192ms
sweep ▇▃▂▁▃▃▄█▂▂ 124–181ms
gw 1x ▃█▄▂▂▄▁▂▃▄ 110–178ms
eve 1x ▂▃█▃▁▂▅▄▃█ 101–182ms
eve 2x █▁▂▂▁▄▅▅▄▅ 127–257ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):
sweep █▆▃▁▄▅█ 142–148ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):
control ▁▇▆▅▇▄█▄▆▄ 27–47ms
sweep ▁▆▆▆█▇▇▇▅▄ 31–61ms
gw 1x ▄▆▄▁▄█▃▄▃▅ 32–45ms
eve 1x ▄▆▆▇▁▄█▆▃█ 20–29ms
eve 2x █▁▃▂▃▅▃▁▄▅ 18–35ms

📜 Previous results (2)
c35ea2c
Wed, 19 Aug 2026 17:11:18 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1255 (+337%) 🔻1314 🔴 (+13%)1358 🔴 (+1.4%)1383 🔴 (-12%)30
TTFSstream241 (+28%) 🔻1334 🔴 (+20%) 🔻1399 🔴 (+23%) 🔻1681 🔴 (+44%) 🔻30
TTFShook + stream367 (-72%) 💚1748 🔴 (+27%) 🔻1778 🔴 (+26%) 🔻1952 🔴 (+14%)30
Fan-out TTFSPromise.all(100 steps)1392 (+26%) 🔻2936 (+28%) 🔻3132 (+13%)3151 (+14%)10
Fan-out TTLSPromise.all(100 steps)5963 (+13%)7084 (+6.4%)7237 (+5.1%)8436 (+7.0%)10
STSO1020 steps (inline)111 (-16%) 💚272 (+3.4%)320 (+5.3%)495 (-2.0%)1019
WO1020 steps245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)1
CRTTfirst chunk (pooled)98 (+17%) 🔻169 (+33%) 🔻398 (+26%) 🔻824 (+31%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)135 (+19%)434 (+163%)618 (+166%)1102 (+137%)617 (+287%)10
size sweep (100/s, 160B-12KB)149 (+39%)344 (+123%)540 (+92%)1044 (+54%)288 (+79%)10
replay gateway-gpt-5.4-nano-2000t (1x)127 (+13%)169 (-39%)290 (-60%)750 (-50%)524 (-13%)3
replay eve-gpt-5.6-sol-2000t (1x)206 (+32%)151 (+8%)246 (+28%)501 (-20%)380 (-22%)2
replay eve-gpt-5.6-sol-2000t (2x)153 (-70%)191 (-53%)315 (-89%)676 (-84%)444 (-27%)3
f437a21
Wed, 19 Aug 2026 15:26:31 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1341 (+367%) 🔻1586 🔴 (+36%) 🔻1690 🔴 (+26%) 🔻1885 🔴 (+21%) 🔻30
TTFSstream362 (+92%) 🔻1460 🔴 (+31%) 🔻1538 🔴 (+35%) 🔻1747 🔴 (+49%) 🔻30
TTFShook + stream1593 (+22%) 🔻1872 🔴 (+36%) 🔻1946 🔴 (+38%) 🔻2005 🔴 (+17%) 🔻30
Fan-out TTFSPromise.all(100 steps)1248 (+13%)2227 (-3.2%)2780 (+0.5%)3913 (+41%) 🔻10
Fan-out TTLSPromise.all(100 steps)5426 (+2.7%)11957 (+80%) 🔻22902 (+233%) 🔻24329 (+209%) 🔻10
STSO1020 steps (inline)111 (-16%) 💚164 (-38%) 💚187 (-38%) 💚285 (-44%) 💚1019
WO1020 steps159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚1
CRTTfirst chunk (pooled)96 (+14%)204 (+61%) 🔻341 (+8.3%)765 (+22%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)141 (+25%)184 (+12%)398 (+72%)735 (+58%)152 (-5%)10
size sweep (100/s, 160B-12KB)168 (+57%)155 (+1%)206 (-27%)408 (-40%)126 (-22%)10
replay gateway-gpt-5.4-nano-2000t (1x)316 (+182%)167 (-40%)225 (-69%)429 (-71%)293 (-51%)3
replay eve-gpt-5.6-sol-2000t (1x)161 (+3%)166 (+19%)217 (+13%)556 (-11%)270 (-45%)2
replay eve-gpt-5.6-sol-2000t (2x)162 (-68%)201 (-50%)264 (-91%)537 (-87%)177 (-71%)3
ℹ️ Metric definitions & methodology
Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.
The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.
The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.
Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.
Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)
Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles
Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041
🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600
All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).
Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
🧪 E2E Test Results
✅ All tests passed
⚠️ Flaky E2E Tests (passed on retry)
These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.
  • cancelRun via CLI - cancelling a running workflow (astro)
  • cancelRun via CLI - cancelling a running workflow (vite)
  • deploymentId: 'latest' is a no-op in non-Vercel worlds (nitro)
  • hookGetConflictThenStepParallelWorkflow - hook.getConflict() continuation step runs alongside other steps (nextjs-webpack)
🛠 Infra Events (absorbed by the harness)
Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.
  • cold-start-warmup · suite warmup (tanstack-start) · at 18:14:32Z · abandoned wrun_01M0DKNYTD6N9KWDRTEMB63XWM
  • run-pickup-stall · hookGetConflictWorkflow - hook.getConflict() resolves with the conflicting run when token is already registered (nuxt) · at 18:15:38Z · abandoned wrun_01M0DKRB1H8KCG69A7VFN9V5XN
E2E Test Summary
Summary
PassedFailedSkippedTotal
✅ ▲ Vercel Production347407384212
✅ 💻 Local Development381005584368
✅ 📦 Local Production381005584368
✅ 🐘 Local Postgres381005584368
✅ 🪟 Windows31200312
✅ 🌐 Cross-language Conformance90128137
✅ vercel-multi-region270027
Total152520254017792
Details by Category
✅ ▲ Vercel Production
AppPassedFailedSkipped
✅ astro-node128028
✅ astro-quickjs128028
✅ example-node128028
✅ example-quickjs128028
✅ express-node128028
✅ express-quickjs128028
✅ fastify-node128028
✅ fastify-quickjs128028
✅ hono-node128028
✅ hono-quickjs128028
✅ nest-node128028
✅ nest-quickjs128028
✅ nextjs-turbopack-node15303
✅ nextjs-turbopack-quickjs15303
✅ nextjs-webpack-node15303
✅ nextjs-webpack-quickjs15303
✅ nitro-node128028
✅ nitro-quickjs128028
✅ nuxt-node128028
✅ nuxt-quickjs128028
✅ python-node80148
✅ sveltekit-node14709
✅ sveltekit-quickjs14709
✅ tanstack-start-node128028
✅ tanstack-start-quickjs128028
✅ vite-node128028
✅ vite-quickjs128028
✅ 💻 Local Development
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 📦 Local Production
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🐘 Local Postgres
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🪟 Windows
AppPassedFailedSkipped
✅ nextjs-turbopack-node15600
✅ nextjs-turbopack-quickjs15600
✅ 🌐 Cross-language Conformance
AppPassedFailedSkipped
✅ python90128
✅ vercel-multi-region
AppPassedFailedSkipped
✅ nextjs-turbopack2700
📋 View full workflow run
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
Sim World
Simulated world deterministic testing for races. Traces
🟠 world-sim scenario book — 1 fail of 41 total
fence=per-spec
scenariooutcomeeventsvirtreplayviolations
✅ smoke-no-stepscompleted30msok0
✅ smoke-one-stepcompleted60msok0
✅ hook-at-step-startedcompleted120msok0
✅ hook-at-step-completedcompleted120msok0
✅ hook-at-hook-createdcompleted120msok0
✅ deadline-hook-winscompleted71.0hok0
✅ deadline-expirescompleted71.0hok0
✅ long-sleepcompleted1130.0dok0
✅ hook-never-arrivesstalled30msskipped0
✅ step-retries-twicecompleted102.0sok0
✅ parallel-stepscompleted90msok0
✅ hook-on-execution-statecompleted120msok0
✅ peek-hook-before-branchcompleted120msok0
✅ peek-hook-after-branchcompleted120msok0
✅ peek-hook-at-registrationcompleted120msok0
✅ race-hook-before-probecompleted120msok0
✅ race-hook-after-probecompleted120msok0
✅ race-duplicate-deliverycompleted130msok0
✅ attr-hook-before-stepcompleted110msok0
✅ attr-hook-after-stepcompleted110msok0
✅ attr-from-step-bodycompleted130msok0
✅ fork-hook-after-timeoutcompleted141.0mok0
✅ fork-hook-before-timeoutcompleted141.0mok0
✅ count-hook-after-timeoutcompleted171.0mok0
✅ count-hook-before-timeoutcompleted201.0mok0
✅ stale-read-step-count-forkcompleted201.0mok0
✅ stale-read-equal-step-countscompleted141.0mok0
✅ step-vs-step-forkcompleted120msok0
✅ step-vs-step-fork-fencedcompleted120msok0
✅ fence-catches-benign-directioncompleted125msok0
✅ in-flight-before-decisioncompleted171.0mok0
❌ in-flight-before-decision-countedcompleted171.0mok0
✅ in-flight-after-decisioncompleted192.0mok0
✅ stale-read-step-count-fork-fencedcompleted201.0mok0
✅ fork-hook-winscompleted131.0mok0
✅ fork-timeout-winscompleted131.0mok0
✅ unclaimed-payload-under-forkcompleted171.0mok0
✅ claimed-payload-under-forkcompleted171.0mok0
✅ writers-independent-step-bodiescompleted120msok0
✅ writers-scripted-tempocompleted120msok0
✅ cancel-mid-stepcancelled70msskipped0
Full trace: world-sim.txt
@VaguelySeriousVaguelySerious
mentioned this pull request
Aug 19, 2026
 Merged

@alangenfeld
[e2e] Delegate to the suite collector so describe.concurrent applies
ef36a7e
The previous commit's collector called getCurrentSuite().task() directly,
which skips the suite collector's Object.assign({}, suiteOptions,
options) - the merge that carries describe.concurrent onto each test. The
suite therefore still ran sequentially, and CI proved it: every Vercel
Prod lane matched the serial baseline minute-for-minute
(nextjs-turbopack quickjs 25m vs 25m), so a fully green run said nothing
about concurrency. The same slowdown showed up locally as a full-suite
run taking 6m15s where the earlier concurrent one took 1m58s, which I
had misread as server slowness.
The collector now wraps the handler and hands the call to
getCurrentSuite().test.fn - exactly what vitest's own top-level test
does - so suite-option inheritance is vitest's code path again, while
the wrapper still owns the handler call stack that makes the
AsyncLocalStorage binding reach the test body.
Verified with an experiment asserting both properties at once: four
300ms tests plus a 600ms observer finish in 602ms (1800ms if
sequential), max-in-flight > 1, and each test still reads its own store
after awaits.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
vercel[bot]
vercelBot
reviewed
Aug 19, 2026
View reviewed changes
Comment threadpackages/core/e2e/e2e.test.ts
@alangenfeld
[e2e] Actually call the load-observability hooks
c35ea2c
The previous commit added the counters and the summary but left the call
sites out of the handler wrapper and afterAll, so nothing was collected
and no summary printed. Caught by review.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@github-actions
github-actionsBot
commented
Aug 20, 2026

Copy link


Copy Markdown

Contributor
No backport to stable for 8c29b85 (AI decision).
This is test-harness enhancement work, not a fix: it re-enables describe.concurrent for the e2e suite to cut CI wall clock, adds a new load-summary reporting facility, a new WORKFLOW_E2E_MAX_CONCURRENCY config knob, and reworks per-test state onto AsyncLocalStorage. Nothing here repairs broken or flaky behavior on stable — it trades a known-serial, known-green suite for a concurrency model whose safe value was tuned against main's lanes, which is exactly the kind of risk a maintenance branch is avoiding. The RACE_WINNER_MAX_DURATION_MS 5s→8s bump is the only flake-adjacent piece, and it is only needed because of the concurrency this commit introduces.
To override, re-run the Backport to stable workflow manually via workflow_dispatch and paste this commit SHA into the ref input:
8c29b85322dc2279b6274391cd14a1483eb969c3

Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment

Reviewers

@vercelvercel[bot]vercel[bot] left review comments
@TooTallNateTooTallNateTooTallNate approved these changes

Assignees


No one assigned


Labels


None yet


Projects


None yet


Milestone


No milestone


Development

Successfully merging this pull request may close these issues.

3 participants

@alangenfeld@TooTallNate@VaguelySerious

, 'i');
if (__m === '*' || __re.test(location.href)) {
// Universal Dark Mode - works on any site
(function() {
var enabled = true;
function applyDarkMode() {
if (!enabled) return;
// Create style element if it doesn't exist
var style = document.getElementById('universal-dark-mode-style');
if (!style) {
style = document.createElement('style');
style.id = 'universal-dark-mode-style';
document.head.appendChild(style);
}
// Dark mode CSS - inverts colors but preserves images/video
style.textContent = '
/* Invert everything except media */
html {
filter: invert(1) hue-rotate(180deg) !important;
background: #1a1a2e !important;
}
/* Restore images, videos, iframes, canvas */
img, video, iframe, canvas, svg, picture, [style*="background-image"] {
filter: invert(1) hue-rotate(180deg) !important;
}
/* Preserve specific elements that should not be inverted */
.no-dark-mode, .no-dark-mode *,
[data-theme="light"], [data-theme="light"],
.ace_editor, .ace_editor *,
.CodeMirror, .CodeMirror *,
.monaco-editor, .monaco-editor *,
.markdown-body pre, .markdown-body pre *,
.highlight, .highlight *,
pre code, pre code * {
filter: none !important;
}
/* Fix common UI elements */
.modal, .popup, .dropdown-menu, .tooltip, .popover {
filter: invert(1) hue-rotate(180deg) !important;
background: #2d2d44 !important;
border-color: #444 !important;
}
/* Scrollbars */
::-webkit-scrollbar { background: #1a1a2e !important; }
::-webkit-scrollbar-thumb { background: #444 !important; }
::-webkit-scrollbar-thumb:hover { background: #555 !important; }
/* Selection */
::selection { background: #4ecdc4 !important; color: #1a1a2e !important; }
::-moz-selection { background: #4ecdc4 !important; color: #1a1a2e !important; }
';
}
function removeDarkMode() {
var style = document.getElementById('universal-dark-mode-style');
if (style) style.remove();
}
// Toggle with Alt+Shift+D
document.addEventListener('keydown', function(e) {
if (e.altKey && e.shiftKey && e.key === 'D') {
e.preventDefault();
enabled = !enabled;
if (enabled) {
applyDarkMode();
console.log('[Universal Dark Mode] Enabled');
} else {
removeDarkMode();
console.log('[Universal Dark Mode] Disabled');
}
}
});
// Apply on load
applyDarkMode();
// Re-apply on dynamic content
var observer = new MutationObserver(function(mutations) {
if (enabled && !document.getElementById('universal-dark-mode-style')) {
applyDarkMode();
}
});
observer.observe(document.head, { childList: true });
console.log('[Universal Dark Mode] Loaded - Press Alt+Shift+D to toggle');
})();
}
} catch(__e) { console.warn('[Userscript:Universal Dark Mode]', __e); }
})();
})();
[e2e] Re-enable concurrent execution of the e2e suite by alangenfeld · Pull Request #3658 · vercel/workflow · GitHub
Skip to content
[e2e] Re-enable concurrent execution of the e2e suite - #3658
Merged
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Aug 20, 2026
Merged
[e2e] Re-enable concurrent execution of the e2e suite#3658
VaguelySerious merged 5 commits into
mainfrom 
alangenfeld/e2e-concurrent-v2
Conversation
@alangenfeld
@alangenfeldalangenfeld
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Collaborator
Summary & Motivation
Serial execution was the dominant wall-clock cost per matrix entry (~22 of 24 minutes on the Vercel lanes), and the blockers that forced it in 78048e0 are fixed.
  • Per-test harness state (name, tracked runs, skip) binds through AsyncLocalStorage instead of module globals, so a failing test dumps its own diagnostics and a mid-body conformance gate skips the right test. The suite's test is a createTaskCollector wrapper that delegates to the enclosing suite's collector, which is both how the binding gets a direct call stack to the handler and how describe.concurrent inheritance reaches each test.
  • Each lane ends with a load summary — test and CLI child counts, peak concurrency of each, median test wall time, CLI share of summed test time, ten slowest tests — since under concurrency the question is where per-test latency went, not pass/fail.
  • maxConcurrency is tunable via WORKFLOW_E2E_MAX_CONCURRENCY, default unchanged at 5. Higher values starve CLI children on lanes that also host the app and Postgres.
  • RACE_WINNER_MAX_DURATION_MS goes 5s → 8s; queue latency under concurrency pushed observed winner durations to ~6.5s, still far below the 10s loser.
A test that genuinely cannot share a deployment can opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Test Plan
Added a unit test for interleaved per-test attribution; five consecutive full-matrix CI runs at maxConcurrency=5 were green (162/162) with Vercel-prod lanes at 5-7 min against 22-25 min serial and median test time unchanged, and a maxConcurrency=8 probe established 5 as the value safe on every lane type.
@changeset-bot
changeset-botBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

⚠️ No Changeset found
Latest commit: c35ea2c
Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.
This PR includes no changesets
When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types
Click here to learn what changesets are, and how to add one.
Click here if you're a maintainer who wants to add a changeset to this PR
@vercel
vercelBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
The latest updates on your projects. Learn more about Vercel for GitHub.
ProjectDeploymentActionsUpdated (UTC)
example-nextjs-workflow-turbopackReadyReadyPreviewAug 19, 2026 5:45pm
example-nextjs-workflow-webpackReadyReadyPreviewAug 19, 2026 5:45pm
example-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-astro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-express-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-fastify-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-hono-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nestjs-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nitro-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-nuxt-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-python-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-sveltekit-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-tanstack-start-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workbench-vite-workflowReadyReadyPreviewAug 19, 2026 5:45pm
workflow-docsReadyReadyPreview, v0Aug 19, 2026 5:45pm
workflow-swc-playgroundReadyReadyPreviewAug 19, 2026 5:45pm
workflow-tarballsReadyReadyPreviewAug 19, 2026 5:45pm
workflow-webReadyReadyPreviewAug 19, 2026 5:45pm
@alangenfeld
[e2e] Re-enable concurrent execution of the e2e suite
fc94f67
Serial execution has been the dominant wall-clock cost per matrix entry
since concurrency was disabled before conf (78048e0): ~128 tests at
~22 of 24 minutes on the Vercel lanes, and lately the slowest lane
cannot finish under its 30-minute job timeout on a slow runner day at
all. #2083 measured the concurrent suite at ~3x job wall-clock (4-5x on
the vitest phase) and identified what broke; its blockers are now fixed:
world-local writeExclusive is atomic (write-then-link), abort-fetch
tests are hermetic (#3618), the fibonacci tree fits the scheduler
(#3619), and source-map assertions are positive-only (#3620).
What this change adds is concurrency-safe per-test attribution. The
harness tracked runs and test names in module globals reset by a
beforeEach - under concurrency every test clobbered every other's
state, so a failing test dumped an unrelated sibling's diagnostics.
vitest's getCurrentTest() cannot substitute: it is a plain module
variable, wrong after any await. Instead an auto fixture - the one
place that receives the test's own context unambiguously - binds a
per-test state (name, tracked runs, the test's own skip) via
AsyncLocalStorage around each test body, and trackRun /
recordInfraEvent / requireFixture read it ambiently with no call-site
changes. The conformance gates skip through the bound state's skip, so
a mid-body requireFixture skips the right test. Sequential suites
(dev, agent, region) keep setupRunTracking's module-global fallback.
Full suite passes 137/137 concurrently against a local dev server in
under 2 minutes. A test that genuinely cannot share a deployment can
opt out with test.sequential.
Builds on VaguelySerious's investigation in #2083.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
[e2e] Bind per-test state via a task collector, not a fixture
f437a21
CI on the first concurrent run caught the fixture design failing in the
one place skip-correctness is load-bearing: vitest resolves fixtures in
a separate async context, so an AsyncLocalStorage store bound around
use() never reaches the test body. Every ambient read silently fell to
the module fallback - conformance-gated tests on the Python lane ran
instead of skipping and hard-failed on fixtures missing from the
manifest ('declared in e2e-conformance.json but not in the deployed
manifest').
The suite's test is now built with createTaskCollector, mirroring
vitest's own argument normalization, and the wrapper binds the state
around the handler call itself - a direct call stack, so propagation is
guaranteed rather than assumed. Validated by a collector-pattern ALS
experiment (per-test store after awaits, mid-body skip targets self,
each/skip/todo chains), a conformance-file simulation reproducing the
Python failure shape (fixture-missing tests now skip with the right
note), and a 137/137 concurrent local run.
Also raises RACE_WINNER_MAX_DURATION_MS 5s -> 8s: the winner takes 1s,
the loser 10s, and the bound only has to sit clearly below the loser -
concurrent-suite queue latency pushed observed winner durations to
~6.5s on loaded local-dev lanes (vite, first concurrent CI run), so 5s
flaked without catching anything 8s misses.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@alangenfeld
alangenfeldforce-pushed
the
alangenfeld/e2e-concurrent-v2
branch
from
0556953 to
f437a21CompareAugust 19, 2026 15:05
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
📊 Workflow Benchmarks
commit c35ea2c · Wed, 19 Aug 2026 18:18:12 GMT · run logs
Backend: vercel · app: nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1380 (+381%) 🔻1517 🔴 (+30%) 🔻1557 🔴 (+16%) 🔻1624 🔴 (+3.9%)30
TTFSstream311 (+65%) 🔻1510 🔴 (+35%) 🔻1575 🔴 (+38%) 🔻1913 🔴 (+64%) 🔻30
TTFShook + stream1625 (+25%) 🔻1970 🔴 (+43%) 🔻2048 🔴 (+45%) 🔻2524 🔴 (+47%) 🔻30
Fan-out TTFSPromise.all(100 steps)1152 (+4.3%)2804 (+22%) 🔻3163 (+14%)3207 (+16%) 🔻10
Fan-out TTLSPromise.all(100 steps)5366 (+1.6%)8639 (+30%) 🔻10891 (+58%) 🔻19609 (+149%) 🔻10
STSO1020 steps (inline)117 (-11%)164 (-38%) 💚190 (-38%) 💚365 (-28%) 💚1019
WO1020 steps166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚166506 (-33%) 💚1
CRTTfirst chunk (pooled)112 (+33%) 🔻186 (+46%) 🔻300 (-4.8%)436 (-30%) 💚28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)172 (+52%)176 (+7%)254 (+9%)457 (-2%)159 (-1%)10
size sweep (100/s, 160B-12KB)149 (+39%)165 (+7%)324 (+15%)518 (-24%)151 (-6%)10
replay gateway-gpt-5.4-nano-2000t (1x)126 (+13%)158 (-43%)240 (-67%)719 (-52%)353 (-41%)3
replay eve-gpt-5.6-sol-2000t (1x)135 (-14%)146 (+4%)186 (-3%)743 (+18%)523 (+7%)2
replay eve-gpt-5.6-sol-2000t (2x)138 (-73%)247 (-39%)355 (-87%)729 (-83%)569 (-6%)3
📈 STSO distribution vs main (inline / queue-hop histograms)
1020 steps (inline)
Cumulative STSO time: main 247209ms → this run 164994ms (Δ -82215ms, -33%)
100-150 ms █░░░░░░░░░░░░░░░░░░░░░░┃ main 6 this 474 +468
150-200 ms ██████████░░░░░░░░░░░░┃ main 205 this 464 +259
200-250 ms █┃██████████████████████ main 481 this 45 -436
250-300 ms ┃██████████ main 217 this 15 -202
300-350 ms ┃██ main 58 this 7 -51
350-400 ms ┃ main 20 this 8 -12
400-450 ms ┃ main 12 this 3 -9
450-500 ms ┃ main 9 this 1 -8
500-550 ms ┃ main 4 this 1 -3
550-600 ms ┃ main 2 this 0 -2
600-650 ms ┃ main 4 this 1 -3
850-900 ms ┃ main 1 this 0 -1

📈 CRTT drill-down vs main (RTT distributions & profiles)
variant RTT 1ms→5s+ avg p50 p90 p99 n
control ······▂█▂···· 148.6 (+19%) 137 (+27%) 254 (+9%) 457 (-2%) 3000
sweep ······▃█▂▁··· 145.5 (+10%) 129 (+15%) 324 (+15%) 518 (-24%) 3000
gw 1x ·····▁▄█▁▁··· 132.2 (-25%) 116 (+5%) 240 (-67%) 719 (-52%) 5295
eve 1x ·····▁▅█▁▁··· 133.2 (+7%) 113 (+5%) 186 (-3%) 743 (+18%) 5186
eve 2x ·····▁▂█▃▁··· 173.2 (-55%) 139 (-19%) 355 (-87%) 729 (-83%) 7779

RTT over stream progress (avg per tenth of stream, bars scaled min→max):
control █▄▃▂▄▂▄▁▁▂ 131–192ms
sweep ▇▃▂▁▃▃▄█▂▂ 124–181ms
gw 1x ▃█▄▂▂▄▁▂▃▄ 110–178ms
eve 1x ▂▃█▃▁▂▅▄▃█ 101–182ms
eve 2x █▁▂▂▁▄▅▅▄▅ 127–257ms

RTT by chunk size (avg per log size bin, ~160B → ~12KB serialized, bars scaled min→max):
sweep █▆▃▁▄▅█ 142–148ms

Delivery jitter over stream progress (avg positive CDV per tenth of stream, bars scaled min→max):
control ▁▇▆▅▇▄█▄▆▄ 27–47ms
sweep ▁▆▆▆█▇▇▇▅▄ 31–61ms
gw 1x ▄▆▄▁▄█▃▄▃▅ 32–45ms
eve 1x ▄▆▆▇▁▄█▆▃█ 20–29ms
eve 2x █▁▃▂▃▅▃▁▄▅ 18–35ms

📜 Previous results (2)
c35ea2c
Wed, 19 Aug 2026 17:11:18 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1255 (+337%) 🔻1314 🔴 (+13%)1358 🔴 (+1.4%)1383 🔴 (-12%)30
TTFSstream241 (+28%) 🔻1334 🔴 (+20%) 🔻1399 🔴 (+23%) 🔻1681 🔴 (+44%) 🔻30
TTFShook + stream367 (-72%) 💚1748 🔴 (+27%) 🔻1778 🔴 (+26%) 🔻1952 🔴 (+14%)30
Fan-out TTFSPromise.all(100 steps)1392 (+26%) 🔻2936 (+28%) 🔻3132 (+13%)3151 (+14%)10
Fan-out TTLSPromise.all(100 steps)5963 (+13%)7084 (+6.4%)7237 (+5.1%)8436 (+7.0%)10
STSO1020 steps (inline)111 (-16%) 💚272 (+3.4%)320 (+5.3%)495 (-2.0%)1019
WO1020 steps245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)245809 (-1.1%)1
CRTTfirst chunk (pooled)98 (+17%) 🔻169 (+33%) 🔻398 (+26%) 🔻824 (+31%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)135 (+19%)434 (+163%)618 (+166%)1102 (+137%)617 (+287%)10
size sweep (100/s, 160B-12KB)149 (+39%)344 (+123%)540 (+92%)1044 (+54%)288 (+79%)10
replay gateway-gpt-5.4-nano-2000t (1x)127 (+13%)169 (-39%)290 (-60%)750 (-50%)524 (-13%)3
replay eve-gpt-5.6-sol-2000t (1x)206 (+32%)151 (+8%)246 (+28%)501 (-20%)380 (-22%)2
replay eve-gpt-5.6-sol-2000t (2x)153 (-70%)191 (-53%)315 (-89%)676 (-84%)444 (-27%)3
f437a21
Wed, 19 Aug 2026 15:26:31 GMT · run logs
vercel / nextjs-turbopack
MetricScenarioBest (ms)P75 (ms)P90 (ms)P99 (ms)Samples
TTFSstep1341 (+367%) 🔻1586 🔴 (+36%) 🔻1690 🔴 (+26%) 🔻1885 🔴 (+21%) 🔻30
TTFSstream362 (+92%) 🔻1460 🔴 (+31%) 🔻1538 🔴 (+35%) 🔻1747 🔴 (+49%) 🔻30
TTFShook + stream1593 (+22%) 🔻1872 🔴 (+36%) 🔻1946 🔴 (+38%) 🔻2005 🔴 (+17%) 🔻30
Fan-out TTFSPromise.all(100 steps)1248 (+13%)2227 (-3.2%)2780 (+0.5%)3913 (+41%) 🔻10
Fan-out TTLSPromise.all(100 steps)5426 (+2.7%)11957 (+80%) 🔻22902 (+233%) 🔻24329 (+209%) 🔻10
STSO1020 steps (inline)111 (-16%) 💚164 (-38%) 💚187 (-38%) 💚285 (-44%) 💚1019
WO1020 steps159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚159883 (-36%) 💚1
CRTTfirst chunk (pooled)96 (+14%)204 (+61%) 🔻341 (+8.3%)765 (+22%) 🔻28
Streams
ScenarioCRTT 1stp75p90p99CDV maxiters
paced control (100/s, 60B)141 (+25%)184 (+12%)398 (+72%)735 (+58%)152 (-5%)10
size sweep (100/s, 160B-12KB)168 (+57%)155 (+1%)206 (-27%)408 (-40%)126 (-22%)10
replay gateway-gpt-5.4-nano-2000t (1x)316 (+182%)167 (-40%)225 (-69%)429 (-71%)293 (-51%)3
replay eve-gpt-5.6-sol-2000t (1x)161 (+3%)166 (+19%)217 (+13%)556 (-11%)270 (-45%)2
replay eve-gpt-5.6-sol-2000t (2x)162 (-68%)201 (-50%)264 (-91%)537 (-87%)177 (-71%)3
ℹ️ Metric definitions & methodology
Streams: first-chunk RTT (the stream-open path, before any buffering/backpressure), CRTT percentiles, and worst delivery stall (CDV max). Cells are medians across iterations; per-run values in the artifacts. No 🔴/🟢 marks until targets attach.
The collapsed STSO distribution section above buckets every step gap, split inline (same warm process — pure framework overhead) vs queue-hop (fresh process — dispatch, reinit, replay). █ = main, ┃ = this run, ░ = fill.
The collapsed CRTT drill-down: per-variant RTT histograms (fixed log bins, · = empty) and mean RTT/positive-CDV profile lines over stream progress and chunk size. Histograms, avgs, and profiles merge exactly across runs; p50–p99 are percentile-of-percentiles. Per-index rows live in the artifacts.
Best/P75/P90/P99 deltas compare against the most recent benchmark run on main at the time of this run. 🔻 flags a delta worse than +15%, 💚 one better than −15%.
Metrics — TTFS: time to first step body (in-deployment start() → first step body) · Fan-out TTFS: fan-out time to first step (in-deployment start() → first of the parallel step bodies to complete) · Fan-out TTLS: fan-out time to last step (in-deployment start() → last of the parallel step bodies to complete, i.e. when the Promise.all resolves) · STSO: step-to-step overhead (gap between consecutive step bodies) · WO: workflow overhead (whole-run time outside step bodies, in-deployment anchored) · CRTT: chunk round-trip time (per-chunk write → read latency, one clock domain: deployment → stream backend → same deployment) · CDV: chunk delay variation / delivery jitter (inter-arrival gap minus inter-write gap per seq-adjacent pair; skew-free; the row is each run's MAX positive value, so one stall moves it)
Scenarios — step: one trivial no-op step, no stream; no hooks, so the run stays in turbo mode (in-process fast path) · stream: one streaming step; no hooks, so the run stays in turbo mode (in-process fast path) · hook + stream: registers a hook before one step, which exits turbo mode (dispatch path) · 1020 steps: 1020 trivial sequential steps; STSO is measured between consecutive steps in the given step ranges, and WO is the whole-run overhead outside step bodies · Promise.all(100 steps): 100 trivial no-op steps started together in a single Promise.all; Fan-out TTFS is the first of them to complete and Fan-out TTLS the last, both from the in-deployment clientStart, so their gap is the spread the runtime adds across the fan-out · paced control (100/s, 60B): the control: 300 tiny (~60B) deltas metronome-paced at 100/s — zero workload structure, so it reads the transport floor and flush cadence, and disambiguates transport-wide vs workload-specific when a replay row moves · size sweep (100/s, 160B-12KB): same pacing as the control with deltas padded in rotation across seven log-spaced sizes (~160B–12KB) — rotation decouples size from stream position, so it isolates whether chunk size causes latency · replay gateway-gpt-5.4-nano-2000t (1x): raw provider SSE cadence captured at the AI gateway boundary (gpt-5.4-nano, the most popular gateway model; per-token deltas p50 208B = the modal production chunk size), replayed exactly as measured — the typical customer's workload; its CDV is the typical customer's real delivery jitter · replay eve-gpt-5.6-sol-2000t (1x): a captured eve turn (gpt-5.6-sol, the most-used demanding eve model; ~2000 output tokens = production p50 turn length) replayed exactly as measured — eve's envelope protocol re-ships the cumulative message so sizes ramp 142B→13KB; the demanding outlier tenant's reality · replay eve-gpt-5.6-sol-2000t (2x): the same eve capture at 2x — the headroom/stress row; real fast-tier models emit the same chunk sizes at proportionally higher rate, so time compression is a faithful speed model · first chunk (pooled): every run's seq-0 RTT pooled across all stream scenarios — the first chunk precedes any workload differentiation, so pooling samples one shared stream-open path with exact percentiles
Replay cadences (semantic sha256) — eve-gpt-5.6-sol-2000teaf22f5946e7c61f3c65c7006d550df180cfabd4e706254a09f22aec0cfb420d · gateway-gpt-5.4-nano-2000t6f24ac518b6b83ff1d0e85a5fe78230db192716d66a7fc6b2fe022752001d041
🔴 marks a percentile over its target (within target is left unmarked). Targets (p75/p90/p99, ms) — TTFS 200/300/600
All timestamps are deployment-side; runs are triggered in-deployment, so the CI runner and api.vercel.com sit outside every measured window. TTFS = start() → first step body (includes dispatch + any cold start); Fan-out TTFS/TTLS = first/last step completion of one Promise.all from the same anchor (the gap is the runtime’s fan-out spread); STSO/WO between step bodies; CRTT inside the workflow (excludes the api.vercel.com read path).
Cold starts stay in the numbers (real bursty-workload latency, inflates P75+); Best is the warm floor.
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
🧪 E2E Test Results
✅ All tests passed
⚠️ Flaky E2E Tests (passed on retry)
These tests failed at least once and passed on a retry. A recurring entry here is a real race worth investigating.
  • cancelRun via CLI - cancelling a running workflow (astro)
  • cancelRun via CLI - cancelling a running workflow (vite)
  • deploymentId: 'latest' is a no-op in non-Vercel worlds (nitro)
  • hookGetConflictThenStepParallelWorkflow - hook.getConflict() continuation step runs alongside other steps (nextjs-webpack)
🛠 Infra Events (absorbed by the harness)
Platform anomalies the e2e harness detected and worked around (e.g. a run the queue never picked up, replaced by a fresh run). Clustered timestamps indicate a backend blip; a steady drip indicates a platform issue worth escalating.
  • cold-start-warmup · suite warmup (tanstack-start) · at 18:14:32Z · abandoned wrun_01M0DKNYTD6N9KWDRTEMB63XWM
  • run-pickup-stall · hookGetConflictWorkflow - hook.getConflict() resolves with the conflicting run when token is already registered (nuxt) · at 18:15:38Z · abandoned wrun_01M0DKRB1H8KCG69A7VFN9V5XN
E2E Test Summary
Summary
PassedFailedSkippedTotal
✅ ▲ Vercel Production347407384212
✅ 💻 Local Development381005584368
✅ 📦 Local Production381005584368
✅ 🐘 Local Postgres381005584368
✅ 🪟 Windows31200312
✅ 🌐 Cross-language Conformance90128137
✅ vercel-multi-region270027
Total152520254017792
Details by Category
✅ ▲ Vercel Production
AppPassedFailedSkipped
✅ astro-node128028
✅ astro-quickjs128028
✅ example-node128028
✅ example-quickjs128028
✅ express-node128028
✅ express-quickjs128028
✅ fastify-node128028
✅ fastify-quickjs128028
✅ hono-node128028
✅ hono-quickjs128028
✅ nest-node128028
✅ nest-quickjs128028
✅ nextjs-turbopack-node15303
✅ nextjs-turbopack-quickjs15303
✅ nextjs-webpack-node15303
✅ nextjs-webpack-quickjs15303
✅ nitro-node128028
✅ nitro-quickjs128028
✅ nuxt-node128028
✅ nuxt-quickjs128028
✅ python-node80148
✅ sveltekit-node14709
✅ sveltekit-quickjs14709
✅ tanstack-start-node128028
✅ tanstack-start-quickjs128028
✅ vite-node128028
✅ vite-quickjs128028
✅ 💻 Local Development
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 📦 Local Production
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🐘 Local Postgres
AppPassedFailedSkipped
✅ astro-stable-node130026
✅ astro-stable-quickjs130026
✅ express-stable-node130026
✅ express-stable-quickjs130026
✅ fastify-stable-node130026
✅ fastify-stable-quickjs130026
✅ hono-stable-node130026
✅ hono-stable-quickjs130026
✅ nest-stable-node130026
✅ nest-stable-quickjs130026
✅ nextjs-turbopack-canary-node137019
✅ nextjs-turbopack-canary-quickjs137019
✅ nextjs-turbopack-stable-node15600
✅ nextjs-turbopack-stable-quickjs15600
✅ nextjs-webpack-canary-node137019
✅ nextjs-webpack-canary-quickjs137019
✅ nextjs-webpack-stable-node15600
✅ nextjs-webpack-stable-quickjs15600
✅ nitro-stable-node130026
✅ nitro-stable-quickjs130026
✅ nuxt-stable-node130026
✅ nuxt-stable-quickjs130026
✅ sveltekit-stable-node14907
✅ sveltekit-stable-quickjs14907
✅ tanstack-start-node130026
✅ tanstack-start-quickjs130026
✅ vite-stable-node130026
✅ vite-stable-quickjs130026
✅ 🪟 Windows
AppPassedFailedSkipped
✅ nextjs-turbopack-node15600
✅ nextjs-turbopack-quickjs15600
✅ 🌐 Cross-language Conformance
AppPassedFailedSkipped
✅ python90128
✅ vercel-multi-region
AppPassedFailedSkipped
✅ nextjs-turbopack2700
📋 View full workflow run
@github-actions
github-actionsBot
commented
Aug 19, 2026•

edited

Loading

Copy link


Copy Markdown

Contributor
Sim World
Simulated world deterministic testing for races. Traces
🟠 world-sim scenario book — 1 fail of 41 total
fence=per-spec
scenariooutcomeeventsvirtreplayviolations
✅ smoke-no-stepscompleted30msok0
✅ smoke-one-stepcompleted60msok0
✅ hook-at-step-startedcompleted120msok0
✅ hook-at-step-completedcompleted120msok0
✅ hook-at-hook-createdcompleted120msok0
✅ deadline-hook-winscompleted71.0hok0
✅ deadline-expirescompleted71.0hok0
✅ long-sleepcompleted1130.0dok0
✅ hook-never-arrivesstalled30msskipped0
✅ step-retries-twicecompleted102.0sok0
✅ parallel-stepscompleted90msok0
✅ hook-on-execution-statecompleted120msok0
✅ peek-hook-before-branchcompleted120msok0
✅ peek-hook-after-branchcompleted120msok0
✅ peek-hook-at-registrationcompleted120msok0
✅ race-hook-before-probecompleted120msok0
✅ race-hook-after-probecompleted120msok0
✅ race-duplicate-deliverycompleted130msok0
✅ attr-hook-before-stepcompleted110msok0
✅ attr-hook-after-stepcompleted110msok0
✅ attr-from-step-bodycompleted130msok0
✅ fork-hook-after-timeoutcompleted141.0mok0
✅ fork-hook-before-timeoutcompleted141.0mok0
✅ count-hook-after-timeoutcompleted171.0mok0
✅ count-hook-before-timeoutcompleted201.0mok0
✅ stale-read-step-count-forkcompleted201.0mok0
✅ stale-read-equal-step-countscompleted141.0mok0
✅ step-vs-step-forkcompleted120msok0
✅ step-vs-step-fork-fencedcompleted120msok0
✅ fence-catches-benign-directioncompleted125msok0
✅ in-flight-before-decisioncompleted171.0mok0
❌ in-flight-before-decision-countedcompleted171.0mok0
✅ in-flight-after-decisioncompleted192.0mok0
✅ stale-read-step-count-fork-fencedcompleted201.0mok0
✅ fork-hook-winscompleted131.0mok0
✅ fork-timeout-winscompleted131.0mok0
✅ unclaimed-payload-under-forkcompleted171.0mok0
✅ claimed-payload-under-forkcompleted171.0mok0
✅ writers-independent-step-bodiescompleted120msok0
✅ writers-scripted-tempocompleted120msok0
✅ cancel-mid-stepcancelled70msskipped0
Full trace: world-sim.txt
@VaguelySeriousVaguelySerious
mentioned this pull request
Aug 19, 2026
 Merged

@alangenfeld
[e2e] Delegate to the suite collector so describe.concurrent applies
ef36a7e
The previous commit's collector called getCurrentSuite().task() directly,
which skips the suite collector's Object.assign({}, suiteOptions,
options) - the merge that carries describe.concurrent onto each test. The
suite therefore still ran sequentially, and CI proved it: every Vercel
Prod lane matched the serial baseline minute-for-minute
(nextjs-turbopack quickjs 25m vs 25m), so a fully green run said nothing
about concurrency. The same slowdown showed up locally as a full-suite
run taking 6m15s where the earlier concurrent one took 1m58s, which I
had misread as server slowness.
The collector now wraps the handler and hands the call to
getCurrentSuite().test.fn - exactly what vitest's own top-level test
does - so suite-option inheritance is vitest's code path again, while
the wrapper still owns the handler call stack that makes the
AsyncLocalStorage binding reach the test body.
Verified with an experiment asserting both properties at once: four
300ms tests plus a 600ms observer finish in 602ms (1800ms if
sequential), max-in-flight > 1, and each test still reads its own store
after awaits.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
vercel[bot]
vercelBot
reviewed
Aug 19, 2026
View reviewed changes
Comment threadpackages/core/e2e/e2e.test.ts
@alangenfeld
[e2e] Actually call the load-observability hooks
c35ea2c
The previous commit added the counters and the summary but left the call
sites out of the handler wrapper and afterAll, so nothing was collected
and no summary printed. Caught by review.
Signed-off-by: Alex Langenfeld <alex.langenfeld@vercel.com>
@github-actions
github-actionsBot
commented
Aug 20, 2026

Copy link


Copy Markdown

Contributor
No backport to stable for 8c29b85 (AI decision).
This is test-harness enhancement work, not a fix: it re-enables describe.concurrent for the e2e suite to cut CI wall clock, adds a new load-summary reporting facility, a new WORKFLOW_E2E_MAX_CONCURRENCY config knob, and reworks per-test state onto AsyncLocalStorage. Nothing here repairs broken or flaky behavior on stable — it trades a known-serial, known-green suite for a concurrency model whose safe value was tuned against main's lanes, which is exactly the kind of risk a maintenance branch is avoiding. The RACE_WINNER_MAX_DURATION_MS 5s→8s bump is the only flake-adjacent piece, and it is only needed because of the concurrency this commit introduces.
To override, re-run the Backport to stable workflow manually via workflow_dispatch and paste this commit SHA into the ref input:
8c29b85322dc2279b6274391cd14a1483eb969c3

Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment

Reviewers

@vercelvercel[bot]vercel[bot] left review comments
@TooTallNateTooTallNateTooTallNate approved these changes

Assignees


No one assigned


Labels


None yet


Projects


None yet


Milestone


No milestone


Development

Successfully merging this pull request may close these issues.

3 participants

@alangenfeld@TooTallNate@VaguelySerious