Edit

Share via

Run event-driven and batch workloads with Azure Functions on Azure Container Apps

Azure Functions on Azure Container Apps combines the productivity of Function-as-a-Service (FaaS) with the flexibility of containerized hosting. This integration allows you to deploy event-driven functions as continuous services while maintaining the ability to handle finite workloads with definitive start and end times.

This platform uses a rich set of triggers and bindings and incorporates advanced Azure Container Apps features, enabling it to execute virtually any containerized workload.

Key benefits

The primary advantages of running Azure Functions on Azure Container Apps include:

  • Unified platform: Run both event-driven and finite workloads on a single platform

  • Flexible hosting: Use containerized environments with advanced deployment capabilities

  • Comprehensive triggers: Support for HTTP, timer, storage, Event Hubs, Cosmos DB, and Service Bus triggers

  • Scalable architecture: Automatic scaling with traffic splitting and revision management

Common use cases

Azure Functions on Azure Container Apps is ideal for event-driven, batch, and API workloads that need rapid scaling, flexible deployment, and seamless integration with Azure services. The following table details the implementation and rationale associated with common scenarios.

Scenario Type Implementation Rationale
Schedule tasks Use Timer triggers that execute code on predefined timed intervals, such as executing data clean-up code or generating reports. Timer triggers reliably ensure code execution at specific, recurring intervals, defining a discrete task timeframe.
Batch or stream processing Use Event Hubs triggers to capture and transform data from IoT or event source streams, or use blob or queue triggers with durable execution patterns (such as fan-out/fan-in) to process large datasets. Functions efficiently process and transform data as soon as events arrive.
Machine learning
(inference/processing)		 Functions can run AI inference by pulling data from a queue or integrating with services using bindings. GPU support is available on Azure Container Apps for compute-intensive machine learning workloads. Functions can wrap the complex logic needed for model processing and leverage highly optimized hardware resources available on Azure Container Apps.
Event-driven workloads
(discrete)		 Use Queue Storage triggers or Service Bus triggers where a message arrival instantly triggers processing. Durable Functions can orchestrate this workload. Functions excel at immediate response to messages and events, managing message queues and processing event streams.
On-demand processing Use HTTP triggers as webhooks or APIs to initiate processing upon request. For asynchronous work, the HTTP trigger can defer the actual work to a queue-triggered function. HTTP endpoints allow manual or programmatic initiation of any workload, providing on-demand execution.
CI/CD runners
(agent execution)		 Although typically containerized tasks, the required trigger logic (such as queue events) can be managed by Functions. The containerized function itself executes the necessary code in response to the event. Functions provide the event processing, scaling, and execution environment necessary to run code triggered by external CI/CD platforms.

Additionally, the following scenarios support critical capabilities:

Event-driven processing

  • Real-time data processing from various Azure services
  • HTTP-triggered APIs and webhooks
  • Timer-based scheduled tasks
  • Message queue processing

Finite and batch workloads

  • Data transformation and ETL processes
  • File processing and validation
  • Scheduled maintenance tasks
  • One-time data migration jobs

Advanced capabilities

The following advanced scenarios feature Azure Functions on Azure Container Apps:

GPU-accelerated workloads

With support for specialized hardware requirements, Azure Functions on Azure Container Apps can run workloads that demand advanced compute capabilities such as:

  • GPU-enabled compute: Serverless GPU resources for AI and machine learning workloads

  • Dedicated workload profiles: High-performance computing for intensive applications

  • Flexible scaling: Scale GPU resources based on demand

Complex stateful workflows

Azure Functions on Azure Container Apps supports advanced workflow management using Durable Functions including:

  • Stateful orchestration: Manage complex, long-running processes

  • Human interaction patterns: Support for approval workflows and user input

  • Monitoring and observability: Built-in tracking for workflow execution

  • Extended processing time: Handle processes that exceed standard limits

Scalable web APIs

You can use advanced features such as custom ingress, traffic management featuring:

  • Custom ingress settings: Advanced traffic routing and load balancing

  • High availability: Built-in redundancy and failover

  • Performance optimization: Automatic scaling based on traffic patterns

Advanced deployment strategies

Azure Functions on Azure Container Apps offers sophisticated deployment management capabilities including:

  • Multi-revision support: Run multiple versions simultaneously

  • Traffic splitting: Gradual rollouts and A/B testing

  • Blue-green deployments: Zero-downtime updates

  • Rollback capabilities: Quick reversion to previous versions

Microservices integration

Azure Functions on Azure Container Apps offers native Dapr integration including:

  • Service invocation: Secure communication between services
  • Pub/Sub messaging: Decoupled event-driven communication
  • State management: Distributed state across services
  • Observability: Built-in monitoring and tracing

Next steps

  • Last updated on