How often ORK polls each peer's /api/status endpoint and writes the result to the status_events cache. Independent of the dashboard cadence below.

How often each card on the status dashboard re-queries ORK's cached state. Updates take effect on the next page load.

Settings for every component's log view in ORK — one configuration governs all of them. Updates take effect on the next log page load.

How often the in-process job poller queries each active peer job for state, and how many consecutive poll failures it tolerates before declaring a job UNREACHABLE (terminal).