Features
Alerts
Get notified when something goes wrong.
Catch failures, backlogs, stuck jobs, and slow queues. Queuedash checks every enabled alert every 5 minutes.
Alert types
| Type | Fires when |
|---|---|
| Failure Rate | At least the threshold % of the finished jobs added in the window (minutes) failed |
| Queue Depth | At least the threshold number of jobs are waiting, delayed, or active (you pick which) |
| Job Stuck | A job has been active for longer than the stuck time (minutes) |
| Processing Time | The average or 95th-percentile run time of jobs added in the window (minutes) reaches the threshold (ms) |
| Job Success Check | No job with the given name, optionally matching a data filter, has succeeded in the window (hours) |
Channels
- Email: one or more addresses, on every plan
- Webhook: a URL, on every plan
- Slack: a channel in your connected Slack workspace, on Pro and Business
Incident lifecycle
- Each alert shows Healthy, Firing, Disabled, or Unknown. Unknown means it hasn't been checked in the last 15 minutes.
- When an alert starts firing, each of its channels gets one notification for the incident, not one per check
- When the condition clears, the channels that were notified get a recovery notification
- The cooldown (15 minutes by default) holds back the notification for a new incident until that long after the previous one. Checks keep running during it.
Create an alert
- Open Alerts in the sidebar and click Create alert
- Pick one or more queues (each gets its own alert) and a type
- Name it and fill in the type's settings. A new Failure Rate alert starts at 10% over a 5-minute window.
- Set the cooldown, choose channels, and click Create
Channels are set up under Alerts → Channels → Add channel.
Recommendations
- Start with Failure Rate and Queue Depth alerts on production queues
- Use Job Success Check for scheduled jobs that must run, like a nightly sync
- Review thresholds weekly after seeing real traffic