docs(crons): Document 7 day max_runtime limit

[wedamija]

Part of a series limiting crons max_runtime to 7 days (10080 minutes), down from 28 days, so in-progress check-ins stop changing after a bounded time.

Document the new limit where max_runtime is defined, and note in the SDK developer docs that monitor upserts clamp larger values.

[vercel]

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

Project	Deployment	Actions	Updated
develop-docs	Ready	Preview	Oct 5, 2026 8:45pm UTC
sentry-docs	Ready	Preview	Oct 5, 2026 8:45pm UTC

[cursor]

The change states a 7 day cap for cron max runtime. The plan checks crons page loads and client errors. An issue escalates when those errors rise above the measured band or crons page loads fall to none.

Rollout complete

Services: sentry-docs, develop-docs.

Plan

What changed

Cron monitor docs now say max_runtime may be at most 10080 minutes. The check-in protocol docs now say Sentry clamps a larger value on monitor upsert. The JavaScript crons include carries the same cap on every guide that embeds it. The change is live after production deploys of sentry-docs and develop-docs.

Risk

A broken include can fail JavaScript crons setup pages that embed javascript-crons-upsert. Those pages share the catch-all route /:path*? with the rest of the site. Wrong copy can lead readers to send a runtime above 7 days. This deploy does not change product clamp behavior.

Intended effect

This change only edits documentation copy. Telemetry cannot show that the 10080 minute wording is live. intended_effect: absent means the published pages still omit the 7 day cap. There is no production counter for that sentence.

Regression watch

A render failure on the catch-all docs route can raise client errors across the site. A missing crons page can drop sampled loads on http.url:*crons* while other pages still load. Path-filtered error events with url:*crons* stayed at 0, so use project error counts as the symptom signal.

Signal	Baseline	Rule	Source
Docs production errors	0-18 events per hour. 140 events in 24h against 11877 pageloads.	Hold in 0-18 per hour. Escalate if the hourly count rises above 18.	Sentry org sentry project docs. Query environment:production on dataset errors. Window 2026-10-04T20:17:14Z to 2026-10-05T20:17:14Z.
Develop-docs production errors	0-12 events per hour. 47 events in 24h.	Hold in 0-12 per hour. Escalate if the hourly count rises above 12.	Sentry org sentry project develop-docs. Query environment:production on dataset errors. Window 2026-10-04T20:17:14Z to 2026-10-05T20:17:14Z.
Crons page loads	0-23 sampled pageload or navigation spans per hour. 137 spans in 24h. Several hours were already 0.	The 24h total should stay near 137. Escalate if that total falls near 0 while docs pageloads stay near 11877.	Sentry org sentry project docs. Query (span.op:pageload OR span.op:navigation) AND http.url:*crons* AND environment:production. Transaction name is /:path*?. Traces sample at 30% for users. Window 2026-10-04T20:17:14Z to 2026-10-05T20:17:14Z.

Not observable

The published 10080 minute wording is not in telemetry. Sampled loads for http.url:*check-ins* on develop-docs were 0 in this window, so that page has no traffic baseline. Error events did not match url:*crons* or url:*check-ins*. Those filters cannot isolate the changed pages.

[cursor]

Verified in production

• Develop-docs errors stayed in band, and the check-ins page still shows the 7 day cap.
• Crons docs still show the 7 day cap, and production errors stayed in band.