Alert rules define what conditions trigger a notification, a PSA ticket, or an OnCall page. Backups comes with default thresholds seeded on setup; you can customize them or create new rules for any condition the system tracks.
The alerts system has two layers: rules (what to watch for and how to respond) and alert events (the log of every time a rule fired).
Built-In Detection
Two background detectors run automatically regardless of your alert rules:
- Missed backup detection — runs every 15 minutes. Flags any device that has not recorded a successful job within its policy's expected interval. Creates a synthetic missed-job record and opens a PSA ticket automatically.
- Failure-streak detection — runs hourly. Flags any device with three or more consecutive failed jobs in the last 48 hours. Opens a PSA ticket and triggers OnCall if configured. De-duplicated to one alert per device per day.
Both write to the alert events log for audit history.
Creating an Alert Rule
- Go to Alerts and click New Rule.
- Enter a name for the rule.
- Pick the Rule Type:
-
failed_job— triggers on any individual job failure -
missed_backup— triggers when a device misses its expected window -
streak— triggers on consecutive failures - Other types (size anomaly, etc.) as available
-
- Set Severity: warning or higher.
- Set the Threshold value and unit (for example, 3 consecutive failures, or 24 hours missed).
- Set Notification Channels — the channels or contacts to notify when the rule fires.
- Tick Create Ticket to open a PSA ticket automatically when this rule fires.
- Tick Trigger OnCall to page the on-call rotation when this rule fires.
- Click Save.
Editing and Disabling Rules
- To edit a rule: open it from the Alerts list and make changes.
- To disable without deleting: uncheck the Enabled toggle on the rule.
- To delete: open the rule and click Delete.
Viewing Alert History
- Go to Alerts → Events (or the Events tab in Alerts).
- Each row in the event log shows: which rule fired, what the subject was (device, job), the message, any linked PSA ticket ID, any linked OnCall incident ID, and whether the alert was resolved.
- Filter by date, rule type, or subject to find specific events.
- Click an event to see its full detail, including the resolved timestamp if it was cleared.
PSA Ticket Integration
When a rule has Create Ticket enabled:
- Backups calls the PSA integration endpoint on each trigger.
- The ticket ID is saved on the alert event and, where applicable, on the job record itself.
- If the PSA call fails, Backups logs a fallback ticket ID so the event record remains traceable.
- To enable PSA integration, go to Settings → Integrations and configure your PSA connection.
OnCall Integration
When a rule has Trigger OnCall enabled:
- Backups sends an incident to the OnCall service with a deduplication key so multiple triggers for the same device do not create duplicate pages.
- The OnCall incident ID is saved on the alert event.
- To enable OnCall integration, go to Settings → Integrations and configure OnCall.
FAQ
The failure-streak alert fired but I only see two failures. Streak detection looks at the last 48 hours. If a third failure occurred and was later cancelled or its record removed, the count may differ from what you expect. Check the full job history for the device to reconstruct the sequence.
A PSA ticket was not created for a failed job.
Confirm the alert rule for failed_job has Create Ticket enabled, and that PSA integration is configured in Settings. Also check the alert events log — if the rule fired, there will be an event record with any error detail.
Can I route different rule types to different PSA queues? Notification channels are configured per rule. Set up separate rules with different channel or queue targets if you need tiered routing.