Alerts notify you when your queues experience issues like high failure rates, large backlogs, or slow processing. Stay informed about problems before they impact users.
Alerts are available on Pro and Enterprise plans. Upgrade your plan to enable alerts. Alert Types bullstudio supports six alert types to cover common queue monitoring scenarios:
Failure Rate Triggers when the percentage of failed jobs exceeds a threshold.
Setting Description Example Threshold Failure rate percentage to trigger alert 10% Time Window Period to calculate failure rate 15 minutes Resolve Threshold Rate to resolve alert 5%
Use case : Detect when your workers are experiencing higher-than-normal error rates.Backlog Exceeded Triggers when the number of waiting jobs exceeds a threshold.
Setting Description Example Threshold Number of waiting jobs to trigger 1000 Resolve Threshold Number to resolve alert 500
Use case : Detect when jobs are piling up faster than workers can process them.Processing Time (Average) Triggers when average job processing time exceeds a threshold.
Setting Description Example Threshold Average time in milliseconds 5000ms Time Window Period to calculate average 10 minutes Resolve Threshold Time to resolve alert 3000ms
Use case : Detect when jobs are taking longer than expected to process.Processing Time (P95) Triggers when the 95th percentile processing time exceeds a threshold.
Setting Description Example Threshold P95 time in milliseconds 10000ms Time Window Period to calculate P95 10 minutes Resolve Threshold Time to resolve alert 7000ms
Use case : Detect outlier slow jobs that may indicate specific issues.Processing Time (P99) Triggers when the 99th percentile processing time exceeds a threshold.
Setting Description Example Threshold P99 time in milliseconds 30000ms Time Window Period to calculate P99 10 minutes Resolve Threshold Time to resolve alert 20000ms
Use case : Catch the slowest jobs that may timeout or cause problems.Missing Workers Triggers when no workers have processed jobs for a specified period.
Setting Description Example Grace Period Time with no activity before alerting 5 minutes
Use case : Detect when all workers have crashed or stopped.Alert Statuses Status Description OK Conditions are within normal thresholds Triggered Threshold exceeded, notification sent
Creating an Alert
Navigate to Alerts
Go to your workspace and click Alerts in the sidebar.
Click Create Alert
Click the Create Alert button.
Configure the alert
Fill in the alert configuration: Field Description Name Descriptive name for the alert Description Optional details about what this alert monitors Connection Which Redis connection to monitor Queue Which queue to monitor (or all queues) Alert Type Select from the available types Configuration Type-specific thresholds Recipients Email addresses for notifications Cooldown Minimum time between notifications
Save
Click Save to create the alert. It will start monitoring immediately.
Alert Configuration Thresholds Set appropriate thresholds based on your normal operating parameters:
Normal failure rate: ~1% → Alert threshold: 5% → Resolve threshold: 2% Normal processing time: ~500ms → Alert threshold: 2000ms → Resolve threshold: 1000ms
Start with higher thresholds and adjust down as you understand your baseline metrics.
Time Windows Time windows determine how data is aggregated:
Shorter windows (5-10 min): More responsive, may be noisyLonger windows (30-60 min): Smoother, may miss brief spikes
Cooldown Period The cooldown prevents alert fatigue:
Minimum: 1 minute
Maximum: 24 hours (1440 minutes)
Default: 15 minutes
During cooldown, no new notifications are sent even if the alert remains triggered.
Recipients Add one or more email addresses to receive notifications:
Notifications sent when alert triggers
Resolution notifications sent when alert returns to OK
Managing Alerts Edit Alert
Click on the alert in the list
Click Edit
Modify settings
Click Save
Enable/Disable Alert Toggle an alert without deleting it:
Click on the alert
Toggle the Enabled switch
Disabled alerts don’t evaluate or send notifications
Delete Alert
Click on the alert
Click Delete
Confirm deletion
Deleting an alert removes all its history. Consider disabling instead if you may need it later.
Test Alert Send a test notification to verify your configuration:
Click on the alert
Click Test
A test notification will be sent to all recipients
Viewing Alert History Each alert maintains a history of triggers:
When the alert was triggered
When it was resolved
Duration of the incident
Use this to:
Identify patterns in issues
Track incident frequency
Measure improvement over time
Alert Best Practices Start with Essential Alerts Begin with these core alerts for most queues:
Failure Rate : Catch bugs and external service issuesBacklog Exceeded : Detect capacity problemsMissing Workers : Know when workers are down
Set Meaningful Thresholds Base thresholds on your actual metrics:
1. Review your dashboard for normal values 2. Set threshold at 2-3x normal 3. Set resolve threshold at 1.5x normal 4. Adjust based on false positives/negatives
Use Appropriate Time Windows Scenario Recommended Window High-volume queues (1000+ jobs/hour) 5-10 minutes Medium-volume queues (100-1000/hour) 15-30 minutes Low-volume queues (<100/hour) 30-60 minutes
Avoid Alert Fatigue
Use resolve thresholds to prevent flapping
Set reasonable cooldown periods
Start with fewer, broader alerts
Refine over time based on actual incidents
Notification Content Alert notifications include:
Alert name and type
Queue and connection name
Current metric value vs threshold
Link to the alert in bullstudio
Timestamp
Next Steps
Setting Up Alerts Guide Step-by-step guide for configuring alerts.
Monitoring Best Practices Learn effective monitoring strategies.
Dashboard View metrics to inform alert thresholds.
