> Markdown version of [/videos/983-reliable-scalability-how-amazon-com-scales-on-aws?t=257](https://www.wearedevelopers.com/videos/983-reliable-scalability-how-amazon-com-scales-on-aws?t=257). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md).

---

# Reliable scalability: How Amazon.com scales on AWS

Amazon's early monolith once bottlenecked crucial deployments. Discover how their pivot to serverless microservices and cell-based isolation now handles 126 million requests per second.

- **Speakers:** [Florian Mair](https://www.wearedevelopers.com/@florian-mair)
- **Event:** World Congress 2024
- **Published:** August 20, 2024
- **Duration:** 22:24
- **URL:** https://www.wearedevelopers.com/videos/983-reliable-scalability-how-amazon-com-scales-on-aws

## Summary

Amazon’s journey to reliable scalability began by dismantling its early monolith—a massive C/C++ binary named Obidos tethered to a single Oracle database. Because this rigid infrastructure bottlenecked simple deployments like one-click buying, the company pivoted to a distributed service-oriented architecture to fulfill its mandate to get big fast. Today, handling massive Prime Day loads, such as DynamoDB processing 126 million requests per second, proves that true scalability is the foundational prerequisite for reliability, ensuring systems perform correctly and consistently regardless of abrupt demand spikes.

The transformation of IMDb illustrates the AWS well-architected framework in action. By migrating from legacy virtual machines to a serverless AWS Lambda backend, IMDb segmented its monolithic workload into distinct microservices seamlessly governed by a federated GraphQL schema. This gateway-based design allows a single client request to concurrently fetch data from highly specific title, rating, and image services without maintaining complex downstream logic. To sustain cultural agility, autonomous two-pizza teams completely own these microservices from initial design to operational deployment, which has drastically minimized on-call friction. Maintaining reliable performance at peaks of over 13,000 requests per second is achieved by utilizing provisioned concurrency to eliminate serverless cold starts, while additionally shielding the infrastructure from malicious traffic with a web application firewall.

For Amazon's physical logistics, Global Ops Robotics manages over 500 sprawling fulfillment centers using aggressive horizontal fault isolation. Implementing a cell-based architecture creates structural bulkheads that strictly isolate localized deployments and database dependencies into discrete AWS accounts. If a critical service crashes within one designated cell, the failure is fiercely contained, protecting the broader global supply chain from cascading outages. By utilizing complex assignment heuristics to evenly distribute fulfillment centers across isolated cells based on scale, this architecture provides a highly resilient blueprint that multi-tenant SaaS platforms can easily emulate to safely quarantine and balance their own customer workloads.

**Keywords:** legacy monolith migration, service-oriented architecture, aws well-architected framework, federated graphql schema, serverless architecture scaling, lambda provisioned concurrency, serverless cold start mitigation, two-pizza microservice teams, on-call incident reduction, application fault isolation, bulkhead architecture patterns, cell-based cloud architecture, blast radius containment, saas multi-tenant resource isolation, web application firewall protection, global edge caching latency

## Chapters

1. **Early monolithic backend architecture at Amazon** (00:03) — How Amazon's early single-binary architecture created deployment bottlenecks that required architectural shifts.
1. **Transitioning from a monolith to service-oriented architecture** (01:50) — How deployment delays in monolithic application binaries drove the extraction of business logic into independent services.
1. **Scaling database infrastructure for massive retail events** (04:17) — How managing millions of concurrent requests during major retail events demonstrates the need for highly scalable database systems.
1. **Defining reliability through the AWS well-architected framework** (05:11) — How managing interconnected service architectures requires applying a recognized reliability framework to ensure consistent system performance.
1. **Modernizing queries using microservices and federated schemas** (06:48) — How replacing monolithic virtual machines with specialized functions and a federated schema streamlines complex data queries.
1. **Improving service ownership with decentralized autonomous teams** (11:19) — How assigning end-to-end operational ownership of individual microservices to autonomous engineering teams drastically reduces operational issues.
1. **Mitigating cold starts in serverless functions using provisioned concurrency** (12:21) — How scaling serverless environments to thousands of requests per second requires provisioned concurrency to eliminate latency spikes.
1. **Protecting application endpoints with web firewalls and edge networks** (14:51) — How combining web application firewalls and global content delivery networks reduces bot traffic and lowers latency.
1. **Structuring warehouse systems for global fulfillment network operations** (16:25) — How operating massive logistics centers requires deploying thousands of localized microservices for inventory management and robotics.
1. **Isolating distributed system failures using cell-based architecture compartments** (18:01) — How isolating regional deployments into separate cloud accounts prevents localized system failures from causing global outages.

## Related Moments

- [Best practices for implementing durable hyperscale serverless systems](https://www.wearedevelopers.com/videos/793-from-0-to-1-000-000-how-to-build-a-serverless-raffle-service-for-hyperscale) (from "From 0 to 1.000.000: How to build a serverless raffle service for hyperscale")
- [Exploring auto-scaling benefits inside event-driven cloud ecosystems](https://www.wearedevelopers.com/videos/590-building-your-way-to-a-serverless-powered-iot-buzzwire-game) (from "Building your way to a serverless powered IOT Buzzwire game")
- [Resolving resource scaling challenges with dynamic serverless deployments](https://www.wearedevelopers.com/videos/1008-serverless-java-in-action-cloud-agnostic-design-patterns-and-tips) (from "Serverless Java in Action: Cloud Agnostic Design Patterns and Tips")
- [Adopting automation and serverless components for efficiency](https://www.wearedevelopers.com/videos/186-effective-java-strategies-and-architectures-for-clouds) (from "Effective Java Strategies and Architectures for Clouds")
- [Transitioning architecture to microservices at Netflix](https://www.wearedevelopers.com/videos/482-building-a-culture-from-chaos) (from "Building a culture from chaos")
- [Solving complex platform architecture challenges at an enterprise scale](https://www.wearedevelopers.com/videos/1209-coffee-with-developers-maria-apazoglou) (from "Coffee with Developers - Maria Apazoglou")

## Related Articles

- [Making Data Warehouses Fast: A Developer’s Story](https://www.wearedevelopers.com/magazine/107-making-data-warehouses-fast-a-developer-s-story)
- [MLops – Deploying, Maintaining And Evolving Machine Learning Models in Production](https://www.wearedevelopers.com/magazine/115-mlops-deploying-maintaining-and-evolving-machine-learning-models-in-production)
- [Why Event-Driven Architecture Isn’t About Speed (and When You Actually Need It)](https://www.wearedevelopers.com/magazine/745-why-event-driven-architecture-isn-t-about-speed-and-when-you-actually-need-it)
- [How we Build The Software of Tomorrow](https://www.wearedevelopers.com/magazine/120-how-we-build-the-software-of-tomorrow)

## Related Jobs

- [Senior Engineer, Infrastructure Platform](https://www.wearedevelopers.com/jobs/ext/328836-senior-engineer-infrastructure-platform) at **Intercom, Inc.**
- [Cloud Foundations Team](https://www.wearedevelopers.com/jobs/ext/1483289-cloud-foundations-team) at **GitHub**
- [Software Engineer L2 - Cloud Infrastructure](https://www.wearedevelopers.com/jobs/ext/1293339-software-engineer-l2-cloud-infrastructure) at **Twilio**
- [Software Engineer L2 - Cloud Infrastructure](https://www.wearedevelopers.com/jobs/ext/1282024-software-engineer-l2-cloud-infrastructure) at **Twilio**
- [Senior Software Engineer, Enterprise Products](https://www.wearedevelopers.com/jobs/ext/1841248-senior-software-engineer-enterprise-products) at **GitHub**
- [Software Engineer L2 - Cloud Infrastructure](https://www.wearedevelopers.com/jobs/ext/1644423-software-engineer-l2-cloud-infrastructure) at **Twilio**