Jan 31 20247 min.

Serverless technologies and patterns on AWS

An overview of serverless technologies, patterns, and associated services on AWS

Full-Stack Cloud Engineer

TABLE OF CONTENTS

Key AWS serverless services
Key tools commonly used for developing and maintaining serverless applications
Key Serverless Patterns
References

It would be beneficial to start by clarifying what serverless is and what it isn't. Serverless can be described as any service where you don't have to manage or allocate underlying hardware resources, allowing it to automatically scale based on demand.

However, there are some deviations from this principle. For instance, in AWS Lambda, there is a default RAM size for a function and a timeout. Setting a RAM size might seem like managing or reserving resources, but the key point is that you stop paying for the resources once the function invocation ends. This illustrates that, although you might specify some parameters due to the nature of the service, the focus is not on the resources themselves.

It's also common to encounter the opinion that serverless equals stateless, which, as you now know, is not accurate. While serverless is by default stateless, the two concepts are not synonymous. Stateless refers to a model where each client request to the server is processed independently, without the server retaining any information from previous interactions.

On the other hand, serverless can indeed be stateful, and some serverless services are inherently stateful. For example, information can be stored during a Lambda invocation in a database like RDS or an object storage like S3. Stateful serverless services, such as AWS Step Functions, can pass user information between invocations.

Another misconception is that any AWS Managed Service is automatically serverless. It's only true that every AWS serverless service is managed, but not the other way around. For example, RDS is a managed service, but only one type of RDS, Aurora Serverless, can be considered both managed and serverless.

Key AWS serverless services

Below, find key serverless services available on AWS:

AWS Lambda

Lambda enables code execution in response to events, automatically managing and scaling compute resources without server provisioning, allowing developers to concentrate on application logic.

AWS Step Functions

Step Functions orchestrates serverless workflows to coordinate AWS services, such as Lambda, enabling efficient app building and updates. It designs event-driven state machines for task execution in sequence or based on conditions, without server management. Automatically scaling with workload changes, it provides a scalable solution for complex business logic execution.

AWS App Runner

App Runner simplifies deployment by automatically deploying and scaling your application based on traffic, freeing developers to focus on coding rather than server management, making it ideal for serverless deployments.

AWS Fargate

Fargate offers a serverless compute engine for containers, eliminating the need for server provisioning or cluster management. Compatible with both ECS and EKS, it automates task execution, scaling, and infrastructure provisioning, simplifying the deployment and management of containerized applications

Amazon DynamoDB

NoSQL database service that eliminates the need for hardware provisioning, setup, and maintenance, offering seamless scalability and automatic scaling of throughput capacity and storage to meet application demands, all without downtime. Its fast and predictable performance, alongside automatic management of database scaling, makes it ideal for serverless architectures, freeing developers from infrastructure management.

Amazon S3

It provides object storage through a web service interface, allowing you to store and retrieve any amount of data from anywhere on the web. With S3, there's no need to manage servers or storage systems; Amazon manages the infrastructure, scalability, durability, and security. S3 is designed to make web-scale computing easier for developers, enabling them to focus on their application logic rather than worrying about the underlying storage infrastructure.

Amazon EventBridge

EventBridge offers a serverless event bus service that seamlessly ingests, processes, and routes events from AWS services, your applications, and SaaS applications, facilitating the creation of event-driven architectures without server or infrastructure management. It automatically scales with event volume, streamlining application integration and workflow automation.

Amazon SQS

Message queuing service for decoupling and scaling microservices, distributed systems, and serverless applications, eliminating the need for server provisioning and queue management. It ensures high availability, processing billions of messages per day, and automatically scales with your application's needs, allowing you to send, store, and receive messages reliably without operational overhead.

Amazon SNS

Pub/sub messaging and mobile notifications service that automates the administration of notification infrastructure, requiring no server management. Capable of processing billions of messages daily without capacity planning, it enables scalable, direct message delivery to users via SMS, email, or other services. This efficiency supports the decoupling of microservices, distributed systems, and serverless applications, facilitating the building of scalable, reliable applications with minimal operational overhead.

AWS AppSync

AppSync simplifies developing GraphQL APIs, managing the infrastructure and operational overhead, and securely connecting to data sources such as DynamoDB and Lambda. It supports real-time data synchronization and enables the creation of collaborative, interactive applications without server management. Its serverless approach allows for automatic scaling with the application's needs, facilitating the efficient development of scalable applications with minimal operational effort.

AWS Amplilfy

Offers a suite of tools and services for building scalable full-stack web and mobile applications on AWS, with features like authentication, API (GraphQL and REST), and storage. It simplifies development, deployment, and hosting processes within a serverless architecture, automatically scaling the backend as needed. This framework allows developers to focus on application code by managing operational overhead, integrating seamlessly with various AWS services, and offering a pay-per-use pricing model for its fully managed services.

Amazon API Gateway

A serverless service that manages the infrastructure for hosting APIs, eliminating the need for server provisioning. It supports REST, HTTP, and WebSocket APIs, integrating seamlessly with AWS services like Lambda, DynamoDB, and S3 for serverless application development. Capable of handling hundreds of thousands of concurrent API calls, it covers traffic management, authorization, monitoring, and API version control. This allows developers to create, publish, and maintain APIs easily, focusing on their core product without the concern of managing infrastructure for API requests.

AWS Sercrets Manager

A centralized solution for managing secrets like database credentials and API keys, with AWS handling infrastructure and scaling. It encrypts secrets at rest using KMS keys and ensures secure transmission over TLS, without requiring server provisioning or management. The service includes features like automatic secret rotation, adhering to security best practices while minimizing manual effort. As a fully managed service, it seamlessly integrates with other AWS services, fitting well within a serverless architecture and allowing developers to concentrate on their applications instead of secret management infrastructure.

Amazon Route 53

Allows you to manage DNS records such as A, AAAA, and CNAME through its console or API, while AWS handles all the necessary infrastructure and scaling for routing internet traffic. This service automatically distributes queries across a global network of DNS resolvers, offering scalability and high availability without the need for direct DNS server management, positioning it as a scalable domain name system (DNS) web service.

Amazon CloudFront

A content delivery network (CDN) service, enables global content distribution through edge locations without server provisioning or management. You can upload content to origins like S3 or EC2 and set caching behaviors via the console or API, while AWS manages the infrastructure and scaling. Features like Lambda@Edge allow custom code execution at edge locations, reducing the reliance on origin servers. As a fully managed service by AWS, CloudFront automatically scales for high traffic, aligning well with serverless architectures for efficient content delivery.

Amazon Cognito

Cognito offers authentication, authorization, and user management for web and mobile applications, scaling automatically to support millions of users without requiring backend infrastructure management. This service simplifies user sign-up, sign-in, and access control, allowing developers to concentrate on application development rather than user authentication and authorization processes.

Key tools commonly used for developing and maintaining serverless applications

X-Ray

AWS X-Ray is a key tool for serverless applications, offering insight into requests across distributed components. It collects tracing data from services like Lambda and API Gateway, creating a service map that illustrates the dependencies and flow of requests, including latency and error rates. X-Ray's automatic error detection simplifies debugging by pinpointing issues in serverless functions. It integrates with various AWS services and offers open-source SDKs for custom components in languages like Node.js and Java. These SDKs facilitate request tracing with minimal configuration changes. Additionally, X-Ray's APIs allow for the creation of custom monitoring and analytics solutions by providing programmatic access to trace data.

CloudFormation

CloudFormation is a key tool for serverless applications due to its ability to model all necessary resources in code, including functions, APIs, and databases, ensuring predictable, repeatable deployments. It integrates with the AWS Serverless Application Model (SAM) to support serverless-specific resources like Lambda and API Gateway. This integration simplifies the deployment process, as CloudFormation automates resource provisioning, trigger setup, and permission management, allowing developers to focus on application logic. With commands like 'aws cloudformation deploy,' serverless applications defined in SAM templates can be deployed, updated, or deleted easily. CloudFormation also supports defining additional resources such as VPC configurations and IAM roles in the same template.

AWS SAM (Serverless Application Model)

It's an extension of CloudFormation. You can do everything you can with CloudFormation and get extra features for local development. The AWS SAM CLI facilitates various aspects of serverless application development on AWS. It enables local testing and debugging by emulating Lambda and API Gateway environments, allowing developers to test serverless applications on their machines before deploying them to the cloud.

With features like local Lambda function invocation and sample payload generation, it helps identify issues early. The CLI streamlines the packaging of application code and dependencies for deployment and manages the upload and deployment process on AWS.

Key Serverless Patterns

Common serverless patterns on AWS encompass a range of architectural designs that utilize AWS's serverless services for efficient, scalable, and cost-effective application development.

Event-Driven Architecture

Triggering functions by events from AWS services like S3, API Gateway, or DynamoDB streams, forming the foundation of serverless architecture. Read more about event-driven architectures on AWS.

Microservices Architecture

Decomposing complex applications into independent microservices that interact through events and serverless messaging queues like Amazon SNS and SQS. Check simplified a simplified overview of event-driven architecture. Read more about design principles for microservices on AWS.