Doctor 预警落地：Linux 状态目录在 SD/eMMC 时先做耐久性加固
问题/场景：状态目录落在 SD/eMMC 容易引发高写放大与寿命风险。前置条件：运行在 Linux，且可调整状态目录或挂载参数。实施步骤：1) 用 doctor 确认风险提示；2) 评估状态目录 I/O 频率；3) 迁移到 SSD/NVMe 或优化挂载；4) 回归检查服务稳定性；5) 记录告警处理 SOP。关键命令：`openclaw doctor --fix`、`openclaw gateway status`。验证方法：doctor 不再提示高风险，运行稳定。风险与边界：低端设备迁移介质可能受硬件限制，需验证。来源归因：GitHub PR #31033 + doctor 流程。
GITHUBDiscovered 2026-03-02Author mzepf
Prerequisites
- Linux host with access to filesystem/mount configuration.
- Service maintenance window available for state directory migration.
Steps
- Run `openclaw doctor --fix` and confirm SD/eMMC warning details.
- Measure current state-path write intensity and identify hot files/directories.
- Move state directory to a more durable disk (SSD/NVMe) or adjust mount strategy.
- Restart gateway and run workload smoke tests for at least one scheduler cycle.
- Document remediation actions as a standard operating checklist.
Commands
openclaw doctor --fixopenclaw gateway statusopenclaw gateway restartVerify
Doctor warning is cleared or downgraded with explicit accepted risk, and gateway stays healthy after restart.
Caveats
- Moving state data without backup can cause irreversible session/history loss.
- Some SBC devices expose eMMC as SSD-like names, device type detection may need manual verification（需验证）.
Source attribution
This tip is aggregated from community/public sources and preserved with attribution.
Open original source ↗