Keeps a run's status file. Every call is cheap and never throws: the file is written
on a job's start, discovery and end, on the run's end, and in between at most every
RUN_STATUS_EVERY_MS; a write that fails goes to onError once, and the sync goes
on without a status — a status must not be what stops a sync.
Keeps a run's status file. Every call is cheap and never throws: the file is written on a job's start, discovery and end, on the run's end, and in between at most every
RUN_STATUS_EVERY_MS; a write that fails goes toonErroronce, and the sync goes on without a status — a status must not be what stops a sync.