KEDA review
A free, cloud-agnostic Kubernetes component for scaling applications and batch jobs from events.
Reviewed by iTechGuides Editors · Editorial team · Updated Oct 2026
KEDA is a free, open-source component deployed in Kubernetes clusters to scale applications and batch jobs in response to events and metrics. It is aimed at teams running Kubernetes workloads that need scaling beyond CPU and memory signals, including workloads that can be inactive until an event arrives. It supports Deployments, StatefulSets, Jobs, and custom resources with a Kubernetes scale subresource, so its scope includes both long-running services and batch processing.
Its main distinction is event-driven scaling alongside the Kubernetes Horizontal Pod Autoscaler. KEDA provides scalers for queues, databases, cloud services, and observability systems, including Apache Kafka, AWS SQS, Azure Service Bus, Google Cloud Pub/Sub, Prometheus, RabbitMQ, Redis, PostgreSQL, MySQL, and MongoDB. CPU and memory triggers and cron schedules are also supported, as is predictive scaling through a PredictKube scaler. Supported workloads can move from zero to one and back to zero. Teams that need a source not covered by a built-in scaler can create external scalers; TriggerAuthentication resources configure credentials for event sources.
KEDA is cloud-agnostic but self-hosted, so teams need to operate it within their Kubernetes environment rather than adopt it as a hosted autoscaling service. The core project is free and open source; commercial support is offered separately by other vendors, while KEDA's support channel is community-based. It suits Kubernetes teams with event-driven scaling requirements and the capacity to manage cluster components. Teams seeking a hosted service or vendor-provided support directly from the core project should consider alternatives.
KEDA pros and cons
- Where it wins
- Scales Deployments, StatefulSets, Jobs, and supported custom resources
- Activates supported workloads from zero and scales them back to zero
- Offers scalers for queues, databases, cloud services, and observability systems
- Where it doesn't
- Requires a self-hosted Kubernetes deployment
- Scaling depends on supported event sources or configured triggers
- Support channels are community-based
KEDA fact sheet, pricing and score →
Advertiser disclosure: iTechGuides is reader-supported. Vendors can pay for top positions in our rankings and for a place on other products' pages, and we may earn a commission when you click some links. Bought places stand at the number bought; this page and our disclosure say so. Paying vendors may get a fuller, more detailed review from our editors, and that review can change their score. How we rank.
Last updated · How we research and update