KEDA 2.4: Event-Driven Autoscaling Evolution

Introduction

KEDA 2.4 — Event-Driven Autoscaling Evolution — was released on September 20, 2021.

This is a practical update aimed at making day‑to‑day Kubernetes work a bit more predictable.

In this release: KEDA 2.4 expands autoscaling capabilities with new scalers, improved HTTP add-on, and enhanced observability for event-driven Kubernetes workloads.

New Scalers & Triggers

• Redis Streams scaler enables autoscaling based on message queue depth in Redis Streams.
• Azure Log Analytics scaler provides scaling based on query results from Azure Log Analytics.
• PostgreSQL scaler supports scaling based on query results from PostgreSQL databases.
• MongoDB scaler enables autoscaling based on MongoDB collection metrics.

HTTP Add-On Improvements

1. Request queue scaling enhancements improve accuracy of scaling decisions based on HTTP request queue depth.
2. Multi-tenant support enables isolated scaling for different applications sharing the same HTTP endpoint.
3. Custom metrics support allows scaling based on application-specific HTTP metrics.
4. Rate limiting integration provides more sophisticated scaling strategies for HTTP workloads.

Observability & Monitoring

• Metrics expansion exposes detailed scaling metrics including trigger values, replica counts, and scaling events.
• Prometheus integration provides comprehensive metrics for dashboards and alerting.
• Event logging improves visibility into scaling decisions and trigger evaluations.
• Health checks ensure KEDA scalers are functioning correctly with detailed status reporting.

Performance & Reliability

• Scaler performance optimizations reduce latency in trigger evaluation and scaling decisions.
• Concurrent scaling improvements enable better handling of multiple ScaledObjects in large clusters.
• Error recovery enhancements provide automatic retry for transient scaler failures.
• Resource efficiency optimizations reduce KEDA operator resource usage.

Operational Enhancements

• Helm chart improvements simplify installation and upgrades with better default values.
• Documentation expansion includes comprehensive guides for all scalers and common use cases.
• Migration tools simplify upgrades from KEDA 1.x and 2.0 versions.
• Community growth demonstrates widespread adoption and active development.

Getting Started

kubectl apply -f https://github.com/kedacore/keda/releases/download/v2.4.0/keda-2.4.0.yaml

Create a ScaledObject:

apiVersion: keda.sh/v1alpha1
kind: ScaledObject
metadata:
  name: redis-scaledobject
spec:
  scaleTargetRef:
    name: worker-deployment
  minReplicaCount: 1
  maxReplicaCount: 10
  triggers:
  - type: redis-streams
    metadata:
      address: redis-service:6379
      stream: my-stream
      consumerGroup: my-consumer-group
      pendingEntriesCount: "5"

Summary

KEDA 2.4 continues to evolve as the standard for event-driven autoscaling, providing teams with powerful tools to scale workloads based on any external event source.