Speed up selfhost e2e: tunable sandbox deadline + 3-way sharding - #1551
Open
RhysSullivan wants to merge 1 commit into
Open
Speed up selfhost e2e: tunable sandbox deadline + 3-way sharding#1551RhysSullivan wants to merge 1 commit into
RhysSullivan wants to merge 1 commit into
Conversation
Deploying with
|
| Status | Name | Latest Commit | Updated (UTC) |
|---|---|---|---|
| ✅ Deployment successful! View logs |
executor-cloud | d03edce | Aug 06 2026, 11:04 PM |
Deploying with
|
| Status | Name | Latest Commit | Preview URL | Updated (UTC) |
|---|---|---|---|---|
| ✅ Deployment successful! View logs |
executor-marketing | d03edce | Commit Preview URL Branch Preview URL |
Aug 06 2026, 11:04 PM |
Contributor
Cloudflare preview
Sign-in is Cloudflare Access (one-time PIN to an allowed email). The preview has its own database and encryption key; it is destroyed when this PR closes. |
@executor-js/cli
@executor-js/config
@executor-js/execution
@executor-js/sdk
@executor-js/codemode-core
@executor-js/runtime-quickjs
@executor-js/plugin-file-secrets
@executor-js/plugin-graphql
@executor-js/plugin-keychain
@executor-js/plugin-mcp
@executor-js/plugin-onepassword
@executor-js/plugin-openapi
executor
commit: |
RhysSullivan
marked this pull request as ready for review
August 7, 2026 02:09
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The selfhost E2E lane is the long pole of CI (~15 min, single serial job). Two changes:
Tunable sandbox deadline. The sandbox-deadline scenario (
resume-after-sandbox-deadline.test.ts) proves approvals granted inside their own windows survive an execution that outlives the sandbox's absolute budget — a ratio, not an absolute duration. It previously waited out the production 5-minute budget in real time (~5.75 min of sleep, ~38% of the lane) and did so only on selfhost: cloud CI's short paused-session window makes it skip there. The selfhost app now acceptsEXECUTOR_SANDBOX_TIMEOUT_MS(validated at boot, refuses malformed values), the e2e boot recipe shrinks it to 20s via a shared constant (e2e/setup/sandbox-timeout.ts, same pattern asexecution-limits.ts), and the scenario scales its delays off the configured budget. It now proves the same race in ~25s on selfhost and still runs the full-length journey on any target without the knob.3-way sharding. The selfhost job now shards like cloud's 4-way matrix. Each shard is its own runner booting its own hermetic instance (own port block + data dir), so the project's shared-bootstrap-admin assumption and
fileParallelism: falsesemantics are unchanged per shard. Artifact upload names already derive fromshard-name.Expected: selfhost E2E wall time drops from ~15 min to ~4–5 min.
Verified locally: the deadline scenario passes in 29s against the 20s budget (and exercises the race — the sleeps genuinely cross the budget); execution-heavy scenarios (
toolkits-mcp,artifact-approval,policy-tool-approval,mcp-execute) pass against the shrunken budget since per-dispatch deadline resets keep normal executions inside it; full--shard=3/3slice passes (27 files, 47 tests); typecheck, lint, format clean.