Alerts & Monitoring

Metoro provides comprehensive alerting and monitoring capabilities to help you stay on top of your Kubernetes infrastructure and applications. You can set up alerts based on various metrics, logs, and traces to get notified when something needs attention.

​

Alert Types

Metoro supports several types of alerts:

1. Metric Alerts: Monitor any metric collected by Metoro, including:

   • CPU and memory usage
   • Network traffic
   • Custom metrics from your applications
   • Container metrics
   • Node metrics

2. Log Alerts: Set up alerts based on log patterns or frequencies:

   • Error frequency
   • Specific log patterns
   • Log volume anomalies
   • Custom log queries

3. Trace Alerts: Monitor your application’s performance:

   • Latency thresholds
   • Error rates
   • Request volume
   • Service dependencies

​

Creating an Alert

To create a new alert:

1. Navigate to the Alerts page
2. Click on the Create Alert button
3. Select the type of alert you want to create
4. Configure your alert conditions

​

Alert Configuration Options

When creating an alert, you can configure:

• Alert Name: A descriptive name for your alert
• Description: Detailed information about what the alert monitors
• Query: The metric, log, or trace query to monitor
• Threshold: The condition that triggers the alert
• Evaluation Period: How often to check the condition
• Alert Destinations: Where to send notifications (Slack, PagerDuty, Webhooks)

​

Alert Destinations

Metoro supports multiple alert destinations:

• Slack: Send alerts to Slack channels
• PagerDuty: Create incidents in PagerDuty
• Webhooks: Send alerts to custom endpoints
• Email: Receive alerts via email

​

Setting Up Alert Destinations

1. Go to Integrations
2. Choose your preferred alert destination
3. Follow the integration setup instructions
4. Use the destination in your alert configurations

​

Monitoring Dashboard

The Alerts Monitor provides a real-time view of your alert status:

Features include:

• Current alert status
• Alert history
• Alert frequency
• Alert resolution tracking
• Alert configuration management

​

Best Practices

1. Start Simple: Begin with basic alerts for critical metrics
2. Avoid Alert Fatigue: Set appropriate thresholds to prevent unnecessary notifications
3. Use Description Fields: Add detailed descriptions to help responders understand and resolve issues
4. Test Your Alerts: Verify alert configurations work as expected
5. Document Response Procedures: Include runbook links in alert descriptions

​

Alert Resolution

When an alert is triggered:

1. The alert status changes to “Firing”
2. Notifications are sent to configured destinations
3. The alert remains active until:
   • The condition returns to normal
   • The alert is manually acknowledged
   • The alert is disabled

​

Managing Alerts

You can manage your alerts through:

1. Alert Rules: View and edit alert configurations
2. Alert History: Track past alert occurrences
3. Alert Status: Monitor current alert states
4. Alert Silencing: Temporarily disable alerts during maintenance

​

Example Alert Configurations

​

CPU Usage Alert

name: "High CPU Usage"
type: "metric"
query: "container_cpu_usage_seconds_total"
threshold: "> 80"
duration: "5m"
severity: "warning"

​

Error Log Alert

name: "High Error Rate"
type: "log"
query: "level:error"
threshold: "> 100"
duration: "15m"
severity: "critical"

​

Latency Alert

name: "High Latency"
type: "trace"
query: "http_server_duration_seconds"
threshold: "> 2"
duration: "5m"
severity: "warning"

​

Troubleshooting Alerts

If you’re experiencing issues with alerts:

1. Check the alert configuration
2. Verify destination integration settings
3. Review alert evaluation logs
4. Confirm metric/log/trace data is being collected
5. Test alert conditions manually

For more information about specific alert destinations, check out:

• Slack Integration
• PagerDuty Integration
• Webhook Integration