Prometheus Metrics Explained: Autoscaling Deployments with KEDA in Kubernetes

What are Prometheus Metrics?

Prometheus metrics are numerical time-series data gathered by the Prometheus monitoring system. These metrics:

• Help track the performance of applications, services, and infrastructure.
• They are stored with timestamps for historical analysis.
• Can be queried using PromQL for flexible, real-time insights.

Types of Prometheus metrics:

1. Counter - Cumulative value that only increases (e.g., total HTTP requests).
2. Gauge - Represents a value that can go up or down (e.g., CPU temperature).
3. Histogram - Samples observations and counts in configurable buckets (e.g., request durations).
4. Summary - Similar to a histogram, but provides quantiles.

What is Prometheus?

Prometheus is an open-source tool used for metrics-based monitoring and alerting. It is a very powerful tool for collecting and querying metric data. It collects data from application services and hosts and stores them in a time-series database. It offers a simple yet powerful data model and a query language (PromQL), and can provide detailed and actionable metrics to analyze an application's performance.

What is KEDA?

KEDA is a lightweight, event-driven autoscaler for Kubernetes that scales workloads based on metrics or events from external systems - including Prometheus metrics.
When you deploy KEDA:

• keda-operator manages scaling logic.
• keda-operator-metrics-apiserver exposes custom metrics to Kubernetes' Horizontal Pod Autoscaler (HPA).

Why Prometheus Metrics Matter

Prometheus metrics are critical for monitoring and scaling modern applications running on Kubernetes. They collect time-series data such as request rates, memory usage, CPU utilization, and custom business metrics, enabling proactive scaling decisions.

When paired with KEDA (Kubernetes Event-Driven Autoscaler), Prometheus metrics can trigger workload scaling dynamically — keeping applications responsive while optimizing resource usage.

Installing KEDA Using Helm

1. Add helm repo

helm repo add kedacore https://kedacore.github.io/charts

2. Update helm repo

helm repo update

3. Install keda helm chart

Using helm2 -

helm install kedacore/keda --namespace keda --version 1.4.2 --name keda

Using helm3

kubectl create namespace keda
helm install keda kedacore/keda --version 1.4.2 --namespace keda

Autoscaling with Prometheus Metrics

Here’s an example ScaledObject configuration for autoscaling using Prometheus metrics:

apiVersion: keda.sh/v1alpha1
kind: ScaledObject
metadata:
  name: prometheus-scaledobject
  namespace: demo3
spec:
  scaleTargetRef:
    apiVersion: argoproj.io/v1alpha1
    kind: Rollout
    name: keda-test-demo3
  triggers:
    - type: prometheus
      metadata:
      serverAddress:  http://<prometheus-host>:9090
      metricName: http_request_total
      query: envoy_cluster_upstream_rq{appId="300", cluster_name="300-0", container="envoy", namespace="demo3", response_code="200" }
      threshold: "50"
  idleReplicaCount: 0                       
  minReplicaCount: 1
  maxReplicaCount: 10

Best Practices for Using Prometheus Metrics with KEDA

• Optimize PromQL queries to avoid overloading Prometheus.
• Set reasonable thresholds to prevent scaling flaps.
• Monitor latency and query execution time.
• Combine Prometheus metrics with business metrics for smarter scaling.

Conclusion

By combining Prometheus metrics with KEDA, Kubernetes workloads can scale intelligently and efficiently. This setup ensures applications remain responsive while optimizing cloud costs. With the right PromQL queries and configurations, autoscaling can become both predictive and cost-effective.

FAQ

Shubham Kumar

Shubham is a DevOps engineer and has worked with multiple organizations. He has worked extensively with cloud technologies including AWS, Kubernetes, and ArgoCD.