Check Alerts

Check Alerts Skill

This skill helps you check and analyze Grafana alerts in the Kagenti platform.

When to Use

• User asks "what alerts are firing?"
• User wants to check alert status
• After platform changes or deployments
• During incident investigation
• When troubleshooting platform issues

What This Skill Does

1. List Firing Alerts: Show all currently active alerts
2. Alert Details: Display alert severity, component, and description
3. Alert History: Check recent alert state changes
4. Query Alert Rules: Verify alert configuration
5. Test Alert Queries: Validate PromQL queries

Examples

Check Firing Alerts

# Get all currently firing alerts from Grafana
kubectl exec -n observability deployment/grafana -- \
  curl -s 'http://localhost:3000/api/alertmanager/grafana/api/v2/alerts' \
  -u admin:admin123 | python3 -c "
import sys, json
alerts = json.load(sys.stdin)
firing = [a for a in alerts if a.get('status', {}).get('state') == 'active']
print(f'Firing alerts: {len(firing)}')
for alert in firing:
    labels = alert.get('labels', {})
    annotations = alert.get('annotations', {})
    print(f\"\\n• {labels.get('alertname')} ({labels.get('severity')})\")
    print(f\"  Component: {labels.get('component')}\")
    print(f\"  Description: {annotations.get('description', 'N/A')[:100]}...\")
"

List All Alert Rules

# Get all configured alert rules
kubectl exec -n observability deployment/grafana -- \
  curl -s 'http://localhost:3000/api/v1/provisioning/alert-rules' \
  -u admin:admin123 | python3 -c "
import sys, json
rules = json.load(sys.stdin)
print(f'Total alert rules: {len(rules)}')
for rule in rules:
    print(f\"  • {rule.get('title')} ({rule.get('labels', {}).get('severity')})\")
"

Check Specific Alert Configuration

# Get configuration for a specific alert
kubectl exec -n observability deployment/grafana -- \
  curl -s 'http://localhost:3000/api/v1/provisioning/alert-rules' \
  -u admin:admin123 | python3 -c "
import sys, json
rules = json.load(sys.stdin)
alert_uid = 'prometheus-down'  # Change this to the alert UID
rule = next((r for r in rules if r.get('uid') == alert_uid), None)
if rule:
    print(f\"Alert: {rule.get('title')}\")
    print(f\"Query: {rule.get('data', [{}])[0].get('model', {}).get('expr')}\")
    print(f\"noDataState: {rule.get('noDataState')}\")
    print(f\"execErrState: {rule.get('execErrState')}\")
"

Test Alert Query Against Prometheus

# Test an alert's PromQL query
QUERY='up{job="kubernetes-pods",app="prometheus"} == 0'

kubectl exec -n observability deployment/grafana -- \
  curl -s -G 'http://prometheus.observability.svc:9090/api/v1/query' \
  --data-urlencode "query=${QUERY}" | python3 -m json.tool

Check Alert Evaluation State

# Check why an alert is firing or not firing
kubectl exec -n observability deployment/grafana -- \
  curl -s 'http://localhost:3000/api/v1/eval/rules' \
  -u admin:admin123 | python3 -m json.tool

Alert Locations in Grafana UI

Access Grafana: https://grafana.localtest.me:9443 Credentials: admin / admin123

Navigation:

1. Alerting → Alert rules - View all configured alerts
2. Alerting → Alert list - See firing/pending alerts
3. Alerting → Silences - Manage alert silences
4. Alerting → Contact points - Check notification settings
5. Alerting → Notification policies - View routing rules

Common Alert Issues

False Positives

• Check noDataState configuration (should be OK for most alerts)
• Verify query matches actual resource type (Deployment vs StatefulSet)
• Test query returns correct results

Alert Not Firing When It Should

• Verify metric exists in Prometheus
• Check alert threshold is appropriate
• Verify for duration isn't too long
• Check noDataState isn't masking the issue

Alert Configuration Not Loading

• Restart Grafana: kubectl rollout restart deployment/grafana -n observability
• Check ConfigMap applied: kubectl get configmap grafana-alerting -n observability
• Verify no YAML syntax errors

Related Documentation

• Alert Runbooks
• Alert Testing Guide
• CLAUDE.md Alert Monitoring
• TODO_INCIDENTS.md - Current incident tracking

Runbooks by Alert

When an alert fires, consult its runbook:

• docs/runbooks/alerts/<alert-uid>.md

Example: If "Prometheus Down" alert fires → docs/runbooks/alerts/prometheus-down.md

🤖 Generated with Claude Code