Metrics & Monitoring

Comprehensive monitoring and observability for your Mend deployment.

Overview

Mend exposes Prometheus metrics for monitoring performance, tracking usage, and alerting on issues.

Metrics Endpoint

GET /metrics

Public No authentication required

curl http://localhost:8080/metrics

Available Metrics

Job Metrics

Metric	Type	Description
mend_jobs_total	Counter	Total jobs processed by type and status
mend_job_duration_seconds	Histogram	Job processing duration
mend_jobs_in_progress	Gauge	Currently processing jobs

Queue Metrics

Metric	Type	Description
mend_queue_depth	Gauge	Current queue depth by type
mend_queue_jobs_enqueued_total	Counter	Total jobs enqueued
mend_queue_jobs_dequeued_total	Counter	Total jobs dequeued

System Metrics

Metric	Type	Description
mend_worker_utilization	Gauge	Worker utilization percentage
mend_memory_usage_bytes	Gauge	Memory usage
mend_disk_usage_bytes	Gauge	Disk usage

Grafana Dashboard

Import the provided Grafana dashboard for visualization:

Job Throughput

• Jobs per second
• Success/failure rates
• Processing times

Queue Health

• Queue depths
• Wait times
• Backlog trends

System Resources

• CPU usage
• Memory usage
• Disk space

Alerting Rules

Example Prometheus alerting rules:

groups:
  - name: mend
    rules:
      - alert: HighQueueDepth
        expr: mend_queue_depth > 100
        for: 5m
        annotations:
          summary: "Queue depth is high"

      - alert: HighFailureRate
        expr: rate(mend_jobs_total{status="failed"}[5m]) > 0.1
        for: 5m
        annotations:
          summary: "Job failure rate is high"

Next Steps

• Set up Health Checks
• Configure Deployment
• Learn about Architecture