AWS re:Invent 2024 - Serverless Compute
Back to all Re:Invent 2024
Table of Contents
AI & ML
 |
This session explores how AWS serverless services can be leveraged to build powerful generative AI applications. The presenters showcase various use cases, including chatbots, document processing, virtual assistants, and content creation, demonstrating how serverless integration with AWS services like Bedrock, Lambda, and Step Functions can enable rapid development and scalable deployment of these AI-powered solutions.
|
Compute
 |
This session provides a comprehensive overview of Amazon ECS, covering its features, benefits, and deployment options. The presenters discuss the security, capacity management, networking, monitoring, and cost optimization capabilities of ECS, highlighting the tradeoffs between the self-managed ECS on EC2 and the fully-managed Fargate modes.
|
 |
The session covers best practices for securing Amazon ECS workloads, including using Amazon Inspector for vulnerability management, AWS Signer for image signing and verification, and Amazon GuardDuty for runtime threat detection. The speakers demonstrate how to integrate these services into a CI/CD pipeline to automate security processes.
|
 |
This talk explores the use of Amazon ECS (Elastic Container Service) for running AI and ML workloads, with a focus on flexibility, reliability, performance, scalability, and observability. The presenters discuss the architectural considerations, compute options, scaling capabilities, and monitoring features of ECS that can help organizations successfully deploy and manage their AI/ML applications.
|
 |
The video explores the trade-offs between using containers and serverless functions for cloud-native applications, focusing on operational aspects, integrations, portability, and pricing. The presenters provide real-world customer examples to illustrate how different organizations have leveraged these technologies to address their specific requirements.
|
 |
This talk explores the use of AWS Fargate, a serverless compute engine, for building event-driven architectures. It discusses various integration patterns, such as event buses, topics, and stores, and how they can be leveraged with Fargate to create scalable, resilient, and loosely coupled applications.
|
 |
The presentation provides a deep dive into the architectural patterns and operational practices used by Amazon to ensure the resilience and availability of the Amazon Elastic Container Service (ECS). It covers key concepts such as regional isolation, availability zone utilization, cellular architectures, and continuous improvement through chaos engineering and a culture of embracing failure.
|
 |
This session discusses best practices for reliable rollouts using Amazon ECS, focusing on the three key dimensions of safety, speed, and visibility. The presenters provide insights into deployment patterns, configuration options, and observability features that can help customers achieve high deployment velocity and confidence in their containerized workloads.
|
Developer Experience
 |
This talk introduces AWS Step Functions Workflow Studio, a visual editor for building state machines. The presenter demonstrates how to use Workflow Studio to create a workflow that processes images, including extracting metadata, creating thumbnails, and storing the results in a DynamoDB table.
|
 |
This session explores how Amazon Q, a conversational AI assistant, can help developers build serverless applications quickly and efficiently. The presenter demonstrates how Q can generate code, write unit tests, and document projects, drastically reducing the time and effort required to develop and deploy applications.
|
 |
The presentation showcases the integration capabilities of Amazon EventBridge Pipes, a service that allows seamless connection of AWS services without the need for custom 'glue code'. The speaker demonstrates how Pipes can be used to filter, enrich, and transform event data, enabling efficient application integration and reducing the operational overhead associated with traditional integration approaches.
|
 |
The video discusses the importance of having an integration strategy to address the challenges faced by modern enterprises, such as the need for speed, data growth, and legacy systems. It presents three customer case studies that showcase how adopting a cloud-based integration strategy, leveraging services like EventBridge and AWS AppSync, can enable organizations to accelerate innovation, improve operational efficiency, and deliver better customer experiences.
|
 |
The video presentation discusses the new private API integration capabilities for Amazon EventBridge and AWS Step Functions, which simplify the integration of private APIs and enable more secure, reliable, and scalable event-driven architectures. The presenters highlight the design tenets of simplicity, security, and reliability that guided the development of this feature, and provide detailed insights into the technical implementation and use cases.
|
 |
The talk discusses the challenges of balancing developer freedom and platform consistency in large organizations, highlighting the importance of platform engineering to enable rapid innovation while maintaining standards. The speakers share insights from their experiences at AWS and Globalization Partners, showcasing strategies and tools like the Well-Architected Framework, Team Topologies, and Backstage to empower developers and drive cloud modernization.
|
 |
The talk discusses architectural principles and patterns for building an event-driven integration platform using AWS EventBridge, emphasizing the importance of event design, schema versioning, and platform engineering practices to enable autonomous application teams while maintaining centralized control and governance.
|
 |
This talk explores the fundamental concepts of integration architecture in distributed systems, including coupling, control flow, message order, and error handling. The speakers provide practical examples and insights on how to build robust and flexible integration patterns using AWS services like SQS, SNS, and EventBridge.
|
 |
This session discusses continuous deployment on Amazon ECS, showcasing how its built-in features enable faster, reliable, and automatic deployment and rollback without affecting application uptime. The presentation covers the release process stages, the role of Amazon ECS in the deployment phase, and demonstrates a live example of handling failed deployments using the deployment circuit breaker functionality.
|
 |
The session explores technical patterns and trade-offs to implement seamless, event-driven front-end experiences using various AWS services. The presenters cover real-time events, long-running processes, and push notifications, providing attendees with a good understanding of these architectures and the associated trade-offs.
|
 |
This talk discusses how to manage event-driven architectures at scale, focusing on enforcing schemas, versioning events, and providing observability. The speakers cover techniques like using API Gateway for schema validation, versioning events with metadata, and leveraging AWS services like EventBridge and CloudWatch to monitor the health and performance of the event-driven system.
|
Observability
 |
This talk explores how to enhance observability in Amazon ECS to gain actionable insights. The presenters demonstrate how to leverage AWS tools and services, such as CloudWatch, Container Insights, and Application Signals, to troubleshoot issues, debug problems, and optimize the cost of running ECS-based applications.
|
Security
 |
This session explores best practices for securing serverless applications, including implementing least privilege, applying defense in depth, leveraging AWS service integrations, and using automation. The presenters cover topics such as input validation, IAM permissions, credential management, and security monitoring to help attendees find the right balance between innovation and security.
|
Serverless
 |
This presentation showcases how to streamline the deployment of web applications on AWS Lambda using the AWS Lambda Web Adapter, a Rust-based runtime extension. It demonstrates a simple approach to package a Next.js application with the Web Adapter and deploy it to Lambda, leveraging features like response streaming and CloudFront integration for improved performance and scalability.
|
 |
The session explores various connectivity options, scaling considerations, and security best practices for building and securing private API workloads using Amazon API Gateway. The presenters also discuss how Itau Unibanco, a large Brazilian bank, has leveraged API Gateway to manage and govern their API ecosystem at scale.
|
 |
This talk explores best practices for optimizing multi-tenant serverless architectures on AWS, focusing on key challenges such as tenant isolation, cost attribution, and scaling integration. The speakers discuss architectural patterns, design considerations, and serverless capabilities that can help address these challenges and enable agile, scalable, and cost-efficient SaaS solutions.
|
 |
This talk provides an overview of techniques to improve throughput and monitoring of serverless streaming workloads on AWS, including parallelizing data processing, reducing processing duration, filtering irrelevant messages, and batch processing. The speaker also covers event source mapping details for Kinesis and Kafka, and introduces new capabilities like provisioned pollers for predictable performance during spiky workloads.
|
 |
This session explores how to build serverless SaaS architectures using AWS services, focusing on key aspects such as tenant isolation, data management, and integration patterns. The presentation covers the technical details and design considerations for transitioning from a hosted solution to a serverless, multi-tenant SaaS model, highlighting the benefits and challenges of this approach.
|
 |
This video discusses the 10-year journey of serverless and containers at AWS, highlighting the core principles of serverless, such as abstracting away infrastructure complexity, composability, and extensible architectures. The presentation also covers recent launches and customer success stories, showcasing how serverless enables faster innovation and agility for businesses.
|
 |
This session covers how to efficiently build, test, and scale serverless applications using Terraform, a popular infrastructure-as-code tool. The presenters discuss best practices, reusable modules, and techniques for managing dependencies and deployments across multiple teams and environments.
|
 |
This session covers the introduction of SnapStart, a new feature in AWS Lambda that reduces cold start times for Python and .NET Lambda functions. The presenters discuss the benefits of SnapStart, how it works, and provide best practices for using it to improve the performance of serverless applications.
|
 |
The session explores how organizations can leverage a combination of Kubernetes and serverless technologies to build modern, event-driven applications. It highlights the challenges of managing heterogeneous compute choices and presents solutions like AWS Controllers for Kubernetes (ACK) and Kube Resource Orchestrator (KRO) to simplify the deployment and integration of serverless resources within a Kubernetes-based architecture.
|
 |
This talk discusses how to simplify multi-tenancy with SaaS applications on AWS Fargate, focusing on four key pillars: availability, security, agility, and pricing. The speaker provides detailed strategies and best practices for building a reliable and cost-effective SaaS platform using AWS services and container technologies.
|
 |
This talk explores the use of AWS Lambda and Apache Kafka for real-time data processing applications. It provides an in-depth look at the architecture and features of these technologies, as well as best practices for managing performance and scaling your data processing pipelines.
|
 |
This session covers best practices for building production-ready serverless applications, including resilience, observability, security, governance, and Lambda function optimization. The speakers provide insights on scaling serverless best practices through service blueprints and architectural patterns, as well as upcoming developments in the serverless space.
|
 |
This session explores advanced workflows with AWS Step Functions, covering topics such as choosing the right workflow type, reducing costs, improving data flow, and managing failures at scale. The speaker demonstrates various patterns and features of Step Functions, including nested workflows, JSONata transformations, and distributed map states, to help attendees build efficient and scalable serverless applications.
|
 |
This session explores how Amazon ECS can be leveraged by platform teams to optimize their application deployment strategies, covering patterns like Account-as-a-Service, Platform-as-a-Service, Cluster-as-a-Service, and Template-as-a-Service, with a case study from Fannie Mae showcasing the benefits of using ECS and serverless containers to power their event-driven loan accounting system.
|