Skip to content
Documentation menu

Job statuses and missing reports

The six states, when a report is due, and when a backup turns Missing.

Last updated

Every backup job in BackupSentinel has one status at a time. It comes from the latest report, or from the absence of one. This page explains each status, when a report is due, and exactly when a backup that has gone quiet is marked Missing.

The six statuses

  • HealthyHealthy: The last report came in and read as a success.
  • WarningWarning: It finished, with skipped files or a partial run.
  • FailedFailed: The report said it failed. Repeats count as a streak.
  • MissingMissing: No report inside the job's window. Silence is not success.
  • WaitingWaiting: Set up, no report yet, so nothing is assumed.
  • PausedPaused: Monitoring is off for the client, by choice.

Healthy, Warning and Failed come from what the report says. Missing comes from what did not arrive. Waiting and Paused are not results: one means there is nothing to judge yet, the other means you turned monitoring off for that client.

Failed reports in a row count as a streak, so a job that failed four times shows "Failed 4 times in a row", not just "Failed".

When a report is due

Each job has an Expected every interval, set when you create it: hourly, every 2, 3, 4, 6, 8 or 12 hours, daily, every 2 days, weekly or monthly. Daily is the default.

The next report is due one interval after the last report was received. A daily job whose report arrived at 02:14 is next due at 02:14 the following day.

  • A new job counts from the moment you created it.
  • If you change the interval, the clock restarts from the last report (or from now, when there has been none), so a job never turns late the moment you edit it.
  • Once a job has at least five gaps between reports, its page suggests an interval based on the typical gap. Take it if it fits your schedule.

The grace window

A report that is a few minutes late is not a failed backup. Each job gets a grace window after its due time, scaled to its interval:

  • Every 8 hours or more often: a quarter of the interval, at least 15 minutes and at most 2 hours.
  • Longer intervals: a tenth of the interval, at least 2 hours.
Grace window and Missing time for each interval
Expected everyGraceMissing after
Hourly15 min1 h 15 min
Every 2 hours30 min2 h 30 min
Every 3 hours45 min3 h 45 min
Every 4 hours1 h5 h
Every 6 hours1 h 30 min7 h 30 min
Every 8 hours2 h10 h
Every 12 hours2 h14 h
Daily2 h 24 min26 h 24 min
Every 2 days4 h 48 min2 d 4 h 48 min
Weekly16 h 48 min7 d 16 h 48 min
Monthly3 d33 d

Between the due time and the end of the grace window the job is late. The Late filter on the jobs list counts it, but it is not Missing yet and nothing is sent.

When a backup turns Missing

BackupSentinel checks every job every 15 minutes. A job whose due time plus grace has passed without a report becomes Missing, and a critical alert opens: "missed its expected window", with the interval it was expected at.

  • If the job already had an open warning, that alert is raised to critical in place, and any acknowledgement or snooze on it is cleared, so someone looks again.
  • The alert goes to your alert channels like any failure, and the job moves to the top of Needs attention.
  • Because the check runs every 15 minutes, a job can turn Missing up to 15 minutes after its grace window ends.

When the next report arrives, it sets the status again. If it says the backup succeeded, the open alert resolves itself and a recovery notice goes to the channels that received the alert.

New jobs: Waiting

A job you have set up but that has not reported yet is Waiting. Nothing about it is assumed: it is neither healthy nor failed. The first report gives it a status.

A Waiting job is held to the same clock as any other. If no report arrives within one interval plus grace of creating it, it becomes Missing. That catches a backup that was set up but never pointed at its address.

Paused clients

Pausing a client turns monitoring off for all of its jobs: until you resume it, for 24 hours, for 7 days or until a date (up to a year ahead).

  • Reports that arrive while a client is paused are still stored, and still update each job's status.
  • No job of a paused client turns Missing, and no alert is sent for it.
  • When the client resumes, the clocks restart from that moment, so its jobs are not flagged for the time they were paused.

Limits

  • Detection works on intervals, not calendars. There is no "weekdays only" schedule: a backup that never runs at weekends will turn Missing after a weekend on a daily interval. Choose the interval that covers its longest normal gap.
  • The due time counts from when the report was received, not from the time the backup ran.
  • The check runs every 15 minutes, so Missing can lag the end of the grace window by up to 15 minutes. A timed pause can also end up to 15 minutes late.