Repository navigation
ci(ios): prepare the runner before the fixture E2E so open does not pay its start - #3251
Conversation
Size Report
Startup median (7 runs, lower is better):
|
![cubic-dev-ai[bot]](https://avatars.githubusercontent.com/in/1082092?s=60&v=4){kind=small}
There was a problem hiding this comment.
All reported issues were addressed across 2 files
Reply with feedback, questions, or to request a fix.
Re-trigger cubic
There was a problem hiding this comment.
All reported issues were addressed across 3 files (changes from recent commits).
Tip: Review your code locally with the cubic CLI to iterate faster.
Re-trigger cubic
|
This PR is ready. I reviewed 1edd8b4 and found nothing that needs to change before merge. All 19 checks pass. The iOS Smoke Tests job ran the changed harness path: the E2E passed with 15 tests, and the uploaded step-history.json shows prepare at 3.4 s and daemon stop at 218 ms. The CI log does not print the takeover notice or adoption diagnostics, so I infer adoption from the 3.4 s prepare time. replays-manual.yml is workflow_dispatch only and did not run on this head, so I checked the removed drag-step clean:daemon calls by reading the code, not by a live run. I did not rerun tests or live devices locally. Not blocking, and you can take or leave these: in live-runner.ts the line No conflicts are known, and nothing else needs to happen before merge. |
thymikee
added
the
ready-for-human
|
* origin/main: (36 commits) fix(android): rebind the test IME when its commit went to a stale input session (callstack#3061) feat: release public workspace packages through npm publish (callstack#3252) fix(snapshot): stop a SwiftUI toolbar host from covering the sheet it encloses (callstack#3162) ci(ios): prepare the runner before the fixture E2E so open does not pay its start (callstack#3251) fix(apple-runner): fence prep spawns behind a start-owned admission (callstack#3239) 0.21.22 test(apple): own the simctl settings plan tests in simctl-settings.test.ts (callstack#3244) fix(limrun): report the session device id in iOS settings refusals (callstack#3243) 0.21.21 feat(remote): add a host-allocated macos-app lease backend (callstack#3236) test(android): bound the screenshot write wait by wall time, not event-loop turns (callstack#3250) feat(recording): cap the touch overlay frame rate at the caller's --fps (callstack#3241) fix(ad-script): let .ad scripts carry scroll --until and wait capture flags (callstack#3197) (callstack#3234) feat(provider-webdriver): keyboard enter, dismiss, and status over WebDriver (callstack#3233) feat(selectors): match role= against snapshot kind with a node-scoped alias window (callstack#3232) fix(provider-webdriver): read field values, placeholders, secure fields, and checked state from page source (callstack#3231) feat(replay): accept --test-ime on test and replay so flow-owned Android opens opt into the test IME (callstack#3235) refactor(daemon): route daemon-level diagnostics through one scope helper (callstack#3242) docs(adr): correct ADR 0031 pointer event delivery evidence (callstack#3245) fix(ios): stop a tap's post-gesture lookup from recording an XCTest failure (callstack#3060) (callstack#3237) ...
1 participant
Summary
The fixture-backed iOS E2E smoke started with a cold runner. The gesture step ended with
pnpm clean:daemon, which stops the daemon and kills its runner by lease. So the first runner-backed command in the E2E paid the full runner start inside its own 90 s request timeout.This failed PR #3243's iOS smoke (job 112216909377). The restarted runner did not connect within 45 s, recovery rebuilt it (49 s), and the deep-link
openneeded the runner for its alert read, so the daemon timed it out at 90 s.Changes:
clean:daemoncalls that discard the prepared runner: after the gesture step inios.yml, and around the drag step inreplays-manual.yml. They are left over from before perf(ios): reuse prepared runner in Settings smoke #2949/perf(ci): reuse iOS runner and narrow native cache key #2952 made later steps reuse the prepared runner. The built-CLI E2E still gets its own daemon, because its code signature differs from the source daemon. A graceful replacement detaches the runner, and the new daemon adopts it.prepare ios-runneras its first step, under the 420 s prepare budget. It runs insidenode-test-tmpdir.ts, so the daemon keeps the run's TMPDIR and claims directory. If adoption is refused, the runner starts again in this step, not inside the firstopen. The harness stops its daemon at the end, before the wrapper deletes those directories.2 workflow files and the E2E harness.
Validation
At
1edd8b4:pnpm check:affected --runandpnpm typecheckpass. At8bc651c(prepare as a workflow step), iOS smoke passed and prepare adopted the preflight runner: 2.0 s, no build, runner uptime about 29 min. The iOS workflow on the new head rechecks it from inside the harness.