2026.2.9 升级后重点做 Cron 可靠性回归验证

场景：社区转发提到 2026.2.9 含“Cron reliability overhaul”。落地时应把“丢触发、重复触发、重启后漏跑”作为三项必测回归，避免只看版本号不看行为。

XDiscovered 2026-02-14Author @yubrew (RT by @openclaw)

Prerequisites

You can run OpenClaw >= 2026.2.9 in a staging environment before production rollout.
At least 2 representative cron jobs are available (one-shot + recurring).

Steps

Clone production-like cron definitions into staging and record expected fire times for a 2-4 hour window.
Upgrade to 2026.2.9+ and run continuous observation; include at least one gateway restart during observation window.
After each trigger, compare `cron runs` timestamps and message delivery timestamps for drift, duplicates, or misses.
Only promote to production when all test jobs pass one-shot + recurring + restart scenarios.

Commands

openclaw gateway status

openclaw gateway restart

openclaw cron list

openclaw cron runs <jobId>

openclaw logs --local-time

Verify

In staged replay, all expected runs fire once each with no duplicate delivery and no restart-induced skip.

Caveats

This source is a community post summary; exact internal fix coverage should be cross-checked with release notes（需验证）.
Network/provider delays can mimic scheduler drift; use both run logs and channel timestamps for judgment.

Source attribution

This tip is aggregated from community/public sources and preserved with attribution.