fix(cron): keep recurring backoff stable after reload by ai-hpc · Pull Request #74068 · openclaw/openclaw

ai-hpc · 2026-04-29T04:33:27Z

Summary

keep recurring cron error backoff on the fixed recurring schedule during reload/recompute
stop applying one-shot cron.retry.backoffMs to recurring jobs after reload
add regression coverage for an errored recurring job with custom one-shot retry backoff

Verification

pnpm test src/cron/service.jobs.test.ts -- --reporter=verbose
git diff --check
pnpm check:changed --base openclaw/main
- passed through core/core-test typecheck
- stopped at lint:core because oxlint config reports: Rule 'no-underscore-dangle' not found in plugin 'eslint'

clawsweeper · 2026-04-29T04:37:41Z

Thanks for the context here. I swept through the related work, and this is now duplicate or superseded.

Keep open for explicit maintainer handling, not cleanup close: this member-authored draft is still draft/conflicting and now conflicts with the current main and release contract that recurring cron backoff honors configured cron.retry.backoffMs.

Root-cause cluster
Relationship: superseded
Canonical: #93051
Summary: Merged recurring-backoff work owns the config-preserving direction on current main and the latest release; this draft overlaps the same cron backoff contract but proposes default-only recompute behavior.

Members:

canonical: fix(cron): honor configured retry.backoffMs for recurring error backoff floor #93051 - Merged PR established that recurring error backoff honors configured cron.retry.backoffMs and added SQLite readback proof.
partial_overlap: fix: retry transient recurring cron failures #87871 - Merged PR introduced recurring transient cron retry behavior and docs that make cron.retry relevant to recurring jobs.
partial_overlap: fix(cron): preserve unresolved next-run backoff #66113 - Merged PR changed the same cron maintenance/recompute backoff boundary.
partial_overlap: fix(cron): add retry policy for one-shot jobs on transient errors (#24355) #24435 - Merged PR introduced the original cron.retry config surface later reused by recurring behavior.

Proposal only: this assessment does not dispatch repair, suppress jobs, mutate sibling items, close, or merge anything.

Canonical path: Close this PR as superseded by #93051.

So I’m closing this here and keeping the remaining discussion on #93051.

Review details

Best possible solution:

Close this PR as superseded by #93051.

Do we have a high-confidence way to reproduce the issue?

Yes for the review finding: source inspection shows the PR removes configured cron.retry.backoffMs from recompute while current main runtime, docs, release, and durable readback coverage preserve it. I did not run tests because this review is read-only.

Is this the best way to solve the issue?

No. The patch changes only recompute and would diverge from current runtime scheduling, docs, and release behavior; the safer path is to preserve configured backoff or make a coordinated maintainer-approved contract change.

Security review:

Security review cleared: The diff changes only cron scheduling logic and a focused test, with no dependency, workflow, secret, permission, package, or artifact execution surface touched.

AGENTS.md: found and applied where relevant.

What I checked:

linked superseding PR: fix(cron): honor configured retry.backoffMs for recurring error backoff floor #93051 (fix(cron): honor configured retry.backoffMs for recurring error backoff floor) is merged at 2026-06-19T19:35:43Z.
cluster evidence: the durable review links that PR in the work cluster or recommended risk path.
no human follow-up: live comments and timeline hydrated by apply contain no non-automation activity after the ClawSweeper review.

Likely related people:

Alix-007: Authored the merged recurring-backoff fix that made recurring error backoff honor configured cron.retry.backoffMs and added durable SQLite readback coverage. (role: recent fix author; confidence: high; commits: d508337637c8, 27ab3e3c2d58, 16fba65cb6ad; files: src/cron/service/timer.ts, src/cron/service/timer.backoff-config-readback.test.ts)
steipete: Authored the merged recurring transient cron retry behavior and docs that extended cron.retry behavior to recurring jobs. (role: adjacent feature author; confidence: high; commits: cd0bc5195389, 89fbe0dc7239, 00c9f811715d; files: src/cron/service/timer.ts, src/cron/service/timer.regression.test.ts, docs/gateway/configuration-reference.md)
mbelinky: Authored merged work on the same cron maintenance/recompute backoff boundary in jobs.ts and timer.ts. (role: adjacent scheduler backoff contributor; confidence: medium; commits: 190a4b48697b; files: src/cron/service/jobs.ts, src/cron/service/timer.ts)
hugenshen: Authored the merged one-shot retry policy and cron.retry config surface that later recurring behavior reused. (role: introduced retry config surface; confidence: medium; commits: ea3955cd78a9; files: src/cron/service/timer.ts, src/config/types.cron.ts, src/config/schema.help.ts)

Codex review notes: model internal, reasoning high; reviewed against 5361e5a0b455.

greptile-apps · 2026-04-29T04:40:28Z

Greptile Summary

This PR adds gateway-level validation of cron expressions for both cron.add and cron.update handlers, rejecting invalid Croner expressions via computeNextRunAtMs before any service mutation occurs. The fix correctly addresses the disabled-job edge case where the absence of next-run computation allowed invalid schedules to be persisted.

The two new tests directly lock in the pre-mutation rejection contract for both the add and disabled-job update paths, complementing the existing post-mutation parse-error test.

Confidence Score: 5/5

Safe to merge — the fix is narrowly scoped to gateway pre-validation and backed by new targeted tests.

No P0 or P1 findings. The validation logic is straightforward, the placement in the handler sequence is correct (after schema/timestamp validation, before delivery assertion and service mutation), nil-schedule is safely short-circuited, and the catch-all in cronScheduleValidationError is acceptable given that every exception computeNextRunAtMs can throw for a cron-kind schedule is genuinely an invalid-request condition.

No files require special attention.

_{Reviews (4): Last reviewed commit: "fix(cron): reject invalid schedule edits..." | Re-trigger Greptile}

greptile-apps · 2026-04-29T04:40:33Z

+function isCronScheduleParseError(err: unknown): err is Error {
+  return (
+    (err instanceof TypeError || err instanceof RangeError) &&
+    (err.message.startsWith("CronPattern:") || err.message.startsWith("CronDate:"))
+  );
+}


Croner error detection tied to message string prefixes

isCronScheduleParseError detects Croner failures by matching the error message against "CronPattern:" and "CronDate:" prefixes. If a future Croner release rewords or restructures its error messages, the defensive catch in addCronJobWithValidation / updateCronJobWithValidation would silently stop working — Croner TypeError/RangeError instances that no longer match the prefix would fall through the throw err branch and revert to the original gateway-crash behavior. Consider adding a comment pinning the expected Croner version (or a snapshot of the exact prefix format) so the fragility is visible during dependency upgrades.

Prompt To Fix With AI

This is a comment left during a code review. Path: src/gateway/server-methods/cron.ts Line: 103-108 Comment: **Croner error detection tied to message string prefixes** `isCronScheduleParseError` detects Croner failures by matching the error message against `"CronPattern:"` and `"CronDate:"` prefixes. If a future Croner release rewords or restructures its error messages, the defensive catch in `addCronJobWithValidation` / `updateCronJobWithValidation` would silently stop working — Croner `TypeError`/`RangeError` instances that no longer match the prefix would fall through the `throw err` branch and revert to the original gateway-crash behavior. Consider adding a comment pinning the expected Croner version (or a snapshot of the exact prefix format) so the fragility is visible during dependency upgrades. How can I resolve this? If you propose a fix, please make it concise.

ai-hpc · 2026-04-29T16:10:10Z

Thanks. I narrowed the PR body to #74459 and kept the fix at the gateway boundary so invalid cron schedules are rejected before cron.update can mutate disabled jobs.

ai-hpc · 2026-04-29T16:13:36Z

@greptile-apps

clawsweeper · 2026-04-29T16:18:23Z

Codex review: needs maintainer review before merge.

What this changes:

The PR imports cron schedule computation into the gateway cron handler, validates cron schedules before cron.add and cron.update, and adds two gateway validation tests proving invalid cron add/update requests do not call the cron service.

Maintainer follow-up before merge:

This is an active contributor PR with a narrow, plausible fix and no clear blocking defect; the next action is maintainer review and targeted validation, especially deciding whether gateway-only validation is sufficient for the linked issue scope.

Review details

Best possible solution:

Land this gateway-boundary validation if maintainers accept the scope, then handle any broader service-level transactional hardening for pre-existing invalid persisted jobs as a separate, explicit follow-up if needed.

Acceptance criteria:

pnpm test src/gateway/server-methods/cron.validation.test.ts -- --reporter=verbose
pnpm check:changed --base origin/main

What I checked:

current_main_mutates_before_validation: On current main, cron.update validates the request envelope and then calls context.cron.update(jobId, patch); the cron service applies applyJobPatch(job, patch) before computing the next run, so invalid disabled-job cron schedules can be written before Croner parsing runs. (src/cron/service/ops.ts:356, 8cf724a381a3)
croner_parse_site: Cron syntax is parsed through computeNextRunAtMs, which resolves cron schedules by constructing new Cron(expr, { timezone, catch: false }); this is the dependency-backed validation path the PR reuses before service mutation. (src/cron/schedule.ts:32, 8cf724a381a3)
pr_adds_pre_service_validation: The PR adds validateCronSchedule / cronScheduleValidationError in the gateway handler and calls it before context.cron.add and context.cron.update, so invalid submitted cron schedules are rejected before the service can mutate state. (src/gateway/server-methods/cron.ts:93, d5e2045a2d92)
pr_test_coverage: The PR adds focused tests for invalid cron.add and disabled-job cron.update schedules, asserting the gateway returns INVALID_REQUEST and does not call the cron service mutation methods. (src/gateway/server-methods/cron.validation.test.ts:325, d5e2045a2d92)
related_issue_context: The linked bug report and maintainer comment in the PR discussion independently describe the disabled-job loophole and support gateway-layer pre-validation as the intended scope for this fix.
security_review: The PR diff touches only src/gateway/server-methods/cron.ts and src/gateway/server-methods/cron.validation.test.ts; it does not alter workflows, dependencies, lockfiles, package metadata, secrets handling, permissions, artifact downloads, or install/build scripts. (d5e2045a2d92)

Likely related people:

yelog: Authored merged cron gateway parse-error handling in fix(cron): catch croner parse errors in cron.add and cron.update handlers #74193, which is the current main behavior this PR builds on and complements. (role: recent adjacent fix author; confidence: high; commits: d2db67e6931b; files: src/gateway/server-methods/cron.ts, src/gateway/server-methods/cron.validation.test.ts)
steipete: Current checkout blame and prior ClawSweeper history place recent cron gateway and scheduler maintenance around the affected handler and schedule evaluation paths with this maintainer. (role: recent maintainer; confidence: medium; commits: 8055e744, d01cb7b65f39, 911cfe2adc27; files: src/gateway/server-methods/cron.ts, src/gateway/server-methods/cron.validation.test.ts, src/cron/schedule.ts)
vincentkoc: Prior related review context links recent cron service operation fixes to this contributor, and the underlying mutation behavior is in src/cron/service/ops.ts and src/cron/service/jobs.ts. (role: adjacent cron service maintainer; confidence: medium; commits: 924271385ba9, 2896107153e5; files: src/cron/service/ops.ts, src/cron/service/jobs.ts)

Remaining risk / open question:

This review was read-only, so I did not run the PR's tests locally; the verdict is based on source, PR diff, provided discussion, prior report context, and dependency call-site inspection.
The PR prevents newly submitted invalid cron schedules from reaching the service, but it does not make CronService.update transactional for externally corrupted or already-persisted invalid cron jobs; maintainers should decide whether that broader repair belongs in this PR or a follow-up.

Codex review notes: model gpt-5.5, reasoning high; reviewed against 8cf724a381a3.

ai-hpc · 2026-04-29T16:24:19Z

Thanks. I agree broader CronService transactional hardening or repair for already-persisted invalid jobs is separate from this PR. This PR intentionally fixes #74459 at the gateway boundary: submitted invalid cron schedules are rejected before cron.add / cron.update can mutate disabled jobs.

Local proof:
pnpm test src/gateway/server-methods/cron.validation.test.ts -- --reporter=verbose passed,
pnpm build passed, and an isolated gateway smoke confirmed the invalid disabled-job edit now returns INVALID_REQUEST while preserving the previous valid schedule in live state and jobs.json.

ai-hpc · 2026-04-29T16:37:10Z

@greptileai

openclaw-barnacle · 2026-06-14T05:05:56Z

This pull request has been automatically marked as stale due to inactivity.
Please add updates or it will be closed.

steipete · 2026-07-12T23:46:31Z

Thanks @ai-hpc. Closing as superseded by the current cron retry contract. Commit 7d76e54f2bbd made reload recomputation preserve configured cron.retry.backoffMs from run end, and #93051 extended and proved that contract through execution and SQLite readback; both are shipped.

This draft removes the configured schedule during recomputation, so it would now regress current runtime and documented behavior rather than fix it. If a distinct reload defect remains, please open a focused current-main issue with before and after timing plus persisted-state evidence.

openclaw-barnacle Bot added gateway Gateway runtime size: M labels Apr 29, 2026

clawsweeper Bot mentioned this pull request Apr 29, 2026

[Bug]: cron add / cron edit with invalid --cron expression logs gateway error and returns UNAVAILABLE instead of INVALID_REQUEST (raw croner TypeError/RangeError leaks to CLI + log) #74066

Closed

greptile-apps Bot reviewed Apr 29, 2026

View reviewed changes

ai-hpc force-pushed the fix/cron-invalid-pattern-errors branch from 650363a to 9503c75 Compare April 29, 2026 05:21

clawsweeper Bot mentioned this pull request Apr 29, 2026

fix(cron): catch croner TypeError/RangeError in add/update and return INVALID_REQUEST #74101

Closed

ai-hpc force-pushed the fix/cron-invalid-pattern-errors branch from 9503c75 to 29c1eb7 Compare April 29, 2026 11:40

ai-hpc marked this pull request as draft April 29, 2026 11:42

ai-hpc force-pushed the fix/cron-invalid-pattern-errors branch from 29c1eb7 to 189784e Compare April 29, 2026 11:52

openclaw-barnacle Bot added size: XS and removed size: M labels Apr 29, 2026

ai-hpc changed the title ~~fix(gateway): reject invalid cron patterns~~ fix(cron): narrow croner parse error handling Apr 29, 2026

clawsweeper Bot mentioned this pull request Apr 29, 2026

[Bug]: cron edit persists invalid --cron expressions on disabled jobs and can leave them enabled after failed enable #74459

Closed

ai-hpc force-pushed the fix/cron-invalid-pattern-errors branch from 8964ed6 to 42a2c40 Compare April 29, 2026 16:01

openclaw-barnacle Bot added size: S and removed size: XS labels Apr 29, 2026

ai-hpc changed the title ~~fix(cron): narrow croner parse error handling~~ fix(cron): reject invalid schedule edits before mutation Apr 29, 2026

This comment was marked as low quality.

Sign in to view

ai-hpc force-pushed the fix/cron-invalid-pattern-errors branch from 42a2c40 to d5e2045 Compare April 29, 2026 16:11

ai-hpc marked this pull request as ready for review April 29, 2026 16:13

ai-hpc force-pushed the fix/cron-invalid-pattern-errors branch from d5e2045 to 4da928c Compare April 29, 2026 20:29

clawsweeper Bot mentioned this pull request Apr 30, 2026

fix: reject invalid cron edits on disabled jobs #74720

Merged

ai-hpc closed this Apr 30, 2026

ai-hpc force-pushed the fix/cron-invalid-pattern-errors branch from 4da928c to 3c9437a Compare April 30, 2026 05:40

openclaw-barnacle Bot added size: XS and removed size: S labels Apr 30, 2026

ai-hpc reopened this Apr 30, 2026

ai-hpc changed the title ~~fix(cron): reject invalid schedule edits before mutation~~ test(cron): assert live disabled schedule rollback Apr 30, 2026

ai-hpc marked this pull request as draft April 30, 2026 05:46

openclaw-barnacle Bot removed the gateway Gateway runtime label Apr 30, 2026

fix(cron): keep recurring backoff stable after reload

32ead20

ai-hpc force-pushed the fix/cron-invalid-pattern-errors branch from e56554d to 32ead20 Compare April 30, 2026 17:32

ai-hpc changed the title ~~test(cron): assert live disabled schedule rollback~~ fix(cron): keep recurring backoff stable after reloa Apr 30, 2026

ai-hpc changed the title ~~fix(cron): keep recurring backoff stable after reloa~~ fix(cron): keep recurring backoff stable after reload Apr 30, 2026

This was referenced May 11, 2026

[Bug] Recurring cron jobs silently skip transient-error retry — only one-shot jobs retry with backoff #46341

Closed

feat(cron): add retry support for recurring jobs without affecting schedule cadence #51001

Closed

clawsweeper Bot mentioned this pull request May 24, 2026

fix(cron): retry transient errors before next natural cron slot (#85888) #85896

Closed

3 tasks

myps6415 mentioned this pull request May 28, 2026

Cron jobs consistently fail with MiniMax 503 overload during early morning (05:00-07:30 CST), manual triggers succeed #85888

Closed

openclaw-barnacle Bot added the stale Marked as stale due to inactivity label Jun 14, 2026

openclaw-barnacle Bot removed the stale Marked as stale due to inactivity label Jun 15, 2026

This was referenced Jun 19, 2026

fix(cron): honor configured retry.backoffMs for recurring error backoff floor #93051

Merged

fix(cron):cron-every-error-backoff-unresolved #95807

Closed

steipete closed this Jul 12, 2026

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

fix(cron): keep recurring backoff stable after reload#74068

fix(cron): keep recurring backoff stable after reload#74068
ai-hpc wants to merge 1 commit into
openclaw:mainfrom
ai-hpc:fix/cron-invalid-pattern-errors

ai-hpc commented Apr 29, 2026 •

edited

Loading

Uh oh!

clawsweeper Bot commented Apr 29, 2026 •

edited

Loading

Uh oh!

greptile-apps Bot commented Apr 29, 2026 •

edited

Loading

Uh oh!

greptile-apps Bot Apr 29, 2026

Uh oh!

This comment was marked as low quality.

ai-hpc commented Apr 29, 2026

Uh oh!

ai-hpc commented Apr 29, 2026

Uh oh!

clawsweeper Bot commented Apr 29, 2026

Uh oh!

ai-hpc commented Apr 29, 2026

Uh oh!

ai-hpc commented Apr 29, 2026

Uh oh!

openclaw-barnacle Bot commented Jun 14, 2026

Uh oh!

steipete commented Jul 12, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Uh oh!

Conversation

ai-hpc commented Apr 29, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Verification

Uh oh!

clawsweeper Bot commented Apr 29, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

greptile-apps Bot commented Apr 29, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Greptile Summary

Confidence Score: 5/5

Uh oh!

greptile-apps Bot Apr 29, 2026

Choose a reason for hiding this comment

Uh oh!

This comment was marked as low quality.

ai-hpc commented Apr 29, 2026

Uh oh!

ai-hpc commented Apr 29, 2026

Uh oh!

clawsweeper Bot commented Apr 29, 2026

Uh oh!

ai-hpc commented Apr 29, 2026

Uh oh!

ai-hpc commented Apr 29, 2026

Uh oh!

openclaw-barnacle Bot commented Jun 14, 2026

Uh oh!

steipete commented Jul 12, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

ai-hpc commented Apr 29, 2026 •

edited

Loading

clawsweeper Bot commented Apr 29, 2026 •

edited

Loading

greptile-apps Bot commented Apr 29, 2026 •

edited

Loading