Serverless Java in Action: Cloud Agnostic Design Patterns and Tips

What if your serverless Java app could scale from zero to 45 pods in seconds? Discover the cloud-agnostic patterns that make it possible.

#1about 2 minutes

The business case for serverless computing

Serverless computing addresses the challenge of matching resources to fluctuating application workloads, which saves money and improves reliability.

#2about 1 minute

Why traditional Java was too slow for serverless

Java's historically long startup and warmup times made it a poor fit for serverless functions compared to lighter languages like JavaScript or Go.

#3about 2 minutes

Using Quarkus and GraalVM for fast Java startup

Modern cloud-native frameworks like Quarkus use native compilation with GraalVM to achieve sub-second startup times and a small memory footprint.

#4about 4 minutes

Writing cloud-agnostic functions to avoid vendor lock-in

Using abstractions like Quarkus Funqy allows developers to write a single serverless function that can be deployed to AWS Lambda or Azure Functions without code changes.

#5about 2 minutes

The hidden vendor lock-in of serverless container platforms

Even when using portable containers, proprietary cloud services for gateways and eventing can still lock your architecture into a single provider.

#6about 4 minutes

Building a portable serverless platform with Knative

Knative extends Kubernetes to provide open-source, cloud-agnostic serverless capabilities, including auto-scaling and eventing with standards like CloudEvents.

#7about 8 minutes

Live demo of an auto-scaling event-driven application

A live demonstration shows a Quarkus application on Knative scaling from zero to over 45 pods in seconds to handle a request burst.

#8about 3 minutes

Key takeaways for building portable serverless Java applications

The main benefit of this approach is the ability to redeploy a serverless application to a different cloud provider in minutes without any code changes.