What is Azure Container Apps? A plain-English explainer
Azure Container Apps is Azure’s serverless container platform: it lets you run a containerised application without provisioning, patching or otherwise managing the cluster underneath it. You give it a container image, and it handles running, scaling — including scaling all the way down to zero instances when there is no traffic — and networking, in the way a serverless service is generally expected to. Under the hood it is built on Kubernetes and KEDA (an event-driven autoscaler), but that infrastructure is deliberately hidden from you; you interact with an application, not a cluster. Because containers are now a mainstream way to package and run software, Container Apps appears in Azure certification study as the “I want containers, but not the operational overhead” option. Here is what it does, when it fits, and how it compares with the alternatives.
What a container is, briefly
A container packages an application together with everything it needs to run — code, dependencies, configuration — so it behaves the same way regardless of where it is deployed. Our what-is-docker explainer covers containers and container images in full; the short version is that a container is a lightweight, portable unit of software, and Kubernetes is the widely used open-source system for running many containers reliably at scale.
Running Kubernetes yourself, however, means managing the cluster it runs on: capacity planning, upgrades, networking and more. Container Apps exists for the case where you want the benefits of running containers without taking on that operational work.
What Container Apps actually does
You supply a container image, and Container Apps runs it as a managed application with a few defining traits:
- Scale to zero — an application with no incoming traffic or events can scale down to no running instances at all, so it costs nothing to sit idle.
- Event-driven and traffic-driven scaling, via KEDA — instances scale up based on HTTP traffic, queue length, or other event sources, not just CPU usage.
- No cluster to manage — Kubernetes and KEDA run underneath, but there is no cluster to provision, patch or upgrade yourself.
- Built-in networking basics — HTTP ingress, internal service-to-service communication between container apps, and revision management for rolling out new versions.
When you would reach for Container Apps
Container Apps fits well for microservices, background processing jobs, and APIs that already exist as containers — or that you would rather package as a container than adapt to a specific serverless-functions programming model. It particularly suits workloads with variable or unpredictable traffic, since scale-to-zero and event-driven scaling mean you are not paying for idle capacity between bursts. It is a poor fit when you specifically need to configure the Kubernetes layer itself — custom networking policies, specific node types, or direct control over the cluster — because that control is exactly what Container Apps abstracts away.
Container Apps versus Azure Kubernetes Service
Azure also offers Azure Kubernetes Service (AKS), a managed Kubernetes service where you still work directly with a cluster, even though Azure handles much of its operation. The distinction is about how much control you want: AKS gives full Kubernetes control — you configure nodes, networking policies and cluster-level settings yourself — while Container Apps gives managed, serverless containers with that layer hidden. Teams that already have deep Kubernetes expertise and need fine-grained cluster control often choose AKS; teams that want to run containers without becoming Kubernetes operators choose Container Apps.
The Azure counterpart to Cloud Run and Fargate
Other clouds offer the same kind of service. Azure Container Apps is broadly Azure’s counterpart to Google Cloud Run and AWS Fargate — all three let you run containers without managing the underlying cluster, and all support scaling down when there is no demand. The specifics of scaling triggers, networking and pricing differ, but the core idea — containers, serverless operation, no cluster to manage — transfers directly between them.
How you pay for it, without the numbers
Container Apps follows the pay-for-use model typical of serverless compute: you are billed for the resources your containers actually consume while running, and an application scaled to zero incurs no compute charge while idle. Exact rates, included allowances and billing granularity change over time, so this article deliberately avoids listing them — check Microsoft’s official Container Apps pricing page for current detail. The concept worth keeping is that scale-to-zero is what makes bursty or intermittent workloads cost-efficient on this model, compared with infrastructure that runs continuously regardless of traffic.
Where Container Apps appears in certification study
Container Apps is relevant to Azure’s administration path as one of several compute options an administrator chooses between. Azure Administrator Associate (AZ-104) covers deploying and managing container-based workloads; Azure Fundamentals (AZ-900) introduces containers and serverless compute conceptually. It is also useful background for the Kubernetes and Cloud Native Associate certification, since understanding what a managed, serverless container platform abstracts away helps clarify what Kubernetes itself is doing underneath a service like AKS.
This article covers what Container Apps is and why it exists — configuring scaling rules, revisions and networking belong in the syllabus itself, which our /revision study library breaks down lesson by lesson for these certifications.
Original practice questions, timed mock exams and revision notes. No card, nothing to pay.