[beekld]

Summary

#489 changed the ERROR log for an unrecoverable HTTP status, such as a 401, from "giving up permanently" to "will not retry until restarted." The synchronizer does stop, but the data system parks it and starts it again after a delay. An operator could read "restarted" as "restart the process," which also resets the park backoff. The message now says that the data system will retry the synchronizer after a delay.

---

Note

Overview
Updates the ERROR log emitted when an HTTP status is treated as unrecoverable by the synchronizer (e.g. most 4xx besides 400/408/429).

The message no longer says the error will not retry until restarted, which could be read as needing a process restart. It now states that the data system will retry it after a delay, matching how the synchronizer is parked and restarted by the data system rather than giving up permanently.

^{Reviewed by Cursor Bugbot for commit 7674d5a. Bugbot is set up for automated code reviews on this repo. Configure here.}