LocalStack: 30k stars, 40M pulls - and just getting started! 🚀

While cloud environments offer great characteristics for production workloads (high scalability, reliability, cost efficiency), the dev experience often falls short due to the centralized, remote execution model of clouds.

Developers are faced with a slow and tedious code → deploy → test → redeploy → … cycle and often find themselves packaging and uploading their serverless applications tens or hundreds of times a day. These inherent limitations and hurdles in developing cloud applications have caused some folks to compare the cloud to the new mainframe era.

“Now, in the 2020s, mainframes are back! They’re just called ‘the cloud’ now, but not much of their essential nature has changed other than the vendor name.” — HackerNews user “jiggawatts”

We have been listening to the pains reported by cloud developers, and we’re determined to empower the dev community so it can focus on what it is best at — developing great products to solve the world’s pressuring technology problems instead of wasting time with inefficient dev&test loops in the cloud.

A brief history of LocalStack

It all started out as a small open source project a couple of years ago, in August 2016. Back then, the main purpose of LocalStack was very simple: We wanted to help our friends and colleagues to turn their commute time on the train into something more productive.

The initial commit (44326584) added support for 8 core AWS APIs (incl. API Gateway, Lambda, DynamoDB, and a few others).

Since then, the project has turned into a flagship project on Github - we’ve recently hit the landmark number of 30k stars on Github, as well as 40Mio pulls of the localstack/localstack image from Docker Hub.

And we are just getting started.

Today, we’re already supporting some ~30 core AWS APIs in our free LocalStack OpenSource project - an achievement that would have never been possible without the great support and contributions from the entire community! ♥️

Additionally, we have recently launched our LocalStack Pro offering, which provides a set of 30+ advanced APIs, additional features, graphical resource browsers, usage dashboards, and much more - to power your local dev productivity and cover even the most sophisticated of your enterprise use cases.

From Chaos engineering, to Local Cloud Pods, to DNS integration - pushing the boundaries of Local Cloud Development

We still love trains, but at some point we realized that we are going beyond the simple use case of supporting our friends to develop while on one of our favorite eco-friendly means of transportation, and that we are actually fundamentally reinventing cloud app development.

We have embarked on a journey to fix the broken cloud software development model and firmly believe a new way of developing and testing cloud applications is possible. As developers and engineers ourselves, we can confidently say:

Yes, we want to be able to fully run and test our serverless apps and cloud based microservices locally.

Yes, we want to take local snapshots of our cloud resources and easily spin them up on a different machine.

Yes, we want to make a local code change and get immediate feedback.

No, we don’t want to constrain teams to keep cloud development costs under control.

And yes - believe it or not - we also want to be able to develop our cloud apps offline!

Today, we are enabling our users to shift their entire cloud stack to their local machine, and this is only the beginning, as we continue to rethink the limits of local development and testing. We are excited to share a few of our recent developments, illustrating the power of your LocalStack:

• Local Cloud Pods allow you to take a persistent snapshot of your LocalStack instance and share it with your team members, enabling collaborative debugging and pre-seeded test environments with shared data and artifacts.
• Local security testing - to support fast and efficient dev&test loops, LocalStack APIs by default do not require credentials, but once your dev cycle approaches the finish line, you can easily leverage the advanced IAM features and policy enforcement for local security testing.
• Transparent execution mode is an approach to running your cloud applications locally and vice versa without any code change whatsoever, by leveraging a local DNS server that automatically resolves all cloud hostnames (e.g., *.amazonaws.com) to a local IP address.
• Chaos engineering is facilitated by systematically injecting faults and errors in the local infrastructure, to test and harden your applications for resilience and fault tolerance.

We are keeping up to speed with supporting your advanced cloud dev use cases - some of the AWS features already supported include:

• Lambda container images are a new mechanism to package and provision Lambda functions, based on pre-built Docker images that implement the Lambda Runtime API.
• Custom CloudFormation Resources spawn user-defined Lambda functions with custom user code to provision CloudFormation resources. Using transparent execution code, existing Custom Resource Lambda functions can be used locally out of the box.
• AppSync resolvers allow you to expose data from DynamoDB/RDS databases or Lambda functions, accessible using GraphQL queries, rendered via Velocity templates.
• Cognito triggers support the full lifecycle of Cognito users and user pools, by running Lambda functions with custom logic for user validation, user migration, or pre-/post-authentication triggers.
• … and much more!

Ecosystem and Partnerships

We continue to focus on strengthening the LocalStack ecosystem and integrations with best-of-breed cloud and Infrastructure as Code (IaC) frameworks, making your local development experience a breeze. We have recently added integrations for CDK, Amplify, Copilot, and Pulumi - complementing the existing integrations for Serverless, Terraform, and AWS CLI.

We are also establishing partnerships to make it even easier for you to integrate LocalStack with your existing environments and dev workflows. For example, you should check out Shipyard, the new way of spinning up ephemeral environments in Kubernetes to turbocharge your application lifecycle. We’ll cover some more exciting news about one-click-deployments of LocalStack on Shipyard in a follow-up blog post soon.

Going multi-cloud and beyond

We’re excited to announce that we are going multi-cloud - focusing on Azure APIs as the next cloud layer, based on all the learnings we have gathered in our journey so far. If you’d like to participate in our Azure preview program, please get in touch with us directly.

But it doesn’t stop there. More and more of our users are asking for LocalStack to support AI apps - we feel the pain of AI engineers that increasingly depend on cloud platforms for their machine learning workflows. Managing the end-to-end lifecycle of AI models (training, deployment, runtime monitoring) involves juggling a huge tech stack, and LocalStack has a track record of simplifying these processes. Also, the increasing demand for AI at the edge has prompted the giants to build tools like AWS IoT Greengrass or Azure IoT Edge, which we look forward to exploring and working on more.

We’re just getting started!

Stay tuned for more updates! We are excited to have the privilege of working with all of you to accelerate your cloud journey and to put developers back in charge. Let’s make cloud development fun again!