/Glossary · AWS
What is AWS Lambda?
AWS Lambda is the AWS serverless compute service that runs code in response to events without you provisioning or managing servers. It scales automatically and charges per request and for execution time based on allocated memory.
How AWS Lambda works
You upload a function as a zip file or container image and connect it to an event source: API Gateway, SQS queues, S3 events, EventBridge, DynamoDB Streams and many others. Lambda creates execution environments on demand, reuses them while traffic keeps coming, and scales them out in parallel as events increase.
Each invocation can run for up to 15 minutes and use up to 10,240 MB of memory; CPU scales in proportion to memory. It supports Node.js, Python, Java, .NET, Ruby and custom runtimes, on x86 or Graviton. SnapStart reduces cold starts for Java, Python and .NET.
When Lambda makes sense
Lambda fits event-driven or variable-traffic workloads: spiky APIs, file processing, system integrations, scheduled jobs and operational automation. If the load is sporadic, paying only per execution is far cheaper than keeping servers running.
With high, steady traffic, processes that run longer than 15 minutes or a need for persistent connections, a service on ECS with Fargate can be more efficient. It is not all or nothing: many architectures use containers for the core and Lambda for events and supporting tasks.
Best practices and common mistakes
Keep functions small and single-purpose, tune memory by measuring with AWS Lambda Power Tuning, use one IAM role per function and set reserved concurrency to protect your databases. For multi-step flows, orchestrate with Step Functions instead of chaining functions that call each other.
Common mistakes: opening a database connection on every invocation without RDS Proxy, ignoring retries and processing duplicate events without idempotency, bundling huge dependencies that lengthen cold starts, and not setting up dead-letter queues for asynchronous failures.
/FAQ
Questions about AWS Lambda
Another question? Write to hello@nubent.com and an engineer will answer.
How much does AWS Lambda cost?
You pay for the number of requests and for duration in GB-seconds based on allocated memory, with a monthly free tier. Functions on Graviton have a lower price per GB-second. Compute Savings Plans also apply to Lambda and lower cost in exchange for a usage commitment.
What is a Lambda cold start?
It is the extra latency when Lambda creates a new execution environment: it downloads the code, starts the runtime and runs your initialization. You reduce it with lean packages, SnapStart on supported runtimes, or provisioned concurrency for latency-sensitive functions.
What is the maximum execution time for Lambda?
An invocation can run for at most 15 minutes. For longer processes, split the work and orchestrate it with Step Functions, or run it as a task on ECS with Fargate or on AWS Batch, which do not have that limit.
How we help with AWS Lambda
Serverless on AWS
Lambda, API Gateway, Step Functions, EventBridge, DynamoDB and SQS: event-driven architectures that scale on their own, tolerate failures and bill only for real usage.
Learn moreAWS Application Development
Backends, APIs, frontends on CloudFront and Amplify, databases and CI/CD: cloud-native software built on AWS and ready to run in production.
Learn moreMCP servers
Custom MCP servers that connect Claude and other agents to your internal systems and AWS, with authentication, least privilege, audit logging and deployment in your account.
Learn moreNext step
What should we build next on AWS?
30 minutes with an engineer. Concrete next steps, whether or not you work with us.