All 7 Lambda & Serverless questions
More in Lambda & Serverless
What is AWS Lambda and how does the serverless computing model differ from traditional server based applications? What event sources can trigger a Lambda function to execute, and how does this support building event driven applications? How does Lambda concurrency work, and what is the difference between reserved concurrency and the account level concurrency limit? How do Lambda layers help share common code and dependencies across multiple functions without duplicating that code in every deployment package? How does Lambda support running functions within a VPC, and what performance considerations does this introduce compared to functions that do not need VPC access? How should an organization approach designing observability and error handling for a complex, multi function serverless architecture built on Lambda?
Advertisement
Back to results for “Lambda & Serverless”
AWS › Lambda & Serverless

What is a cold start in Lambda, and what strategies help reduce its impact on application performance?

Intermediate

Answer

A cold start occurs when Lambda needs to initialize a brand new execution environment for a function that has not run recently, adding some additional latency compared to a warm invocation that reuses an already initialized environment, and strategies to reduce cold start impact include choosing a lighter weight runtime, minimizing the size of your deployment package and its dependencies, and using Provisioned Concurrency to keep a specified number of execution environments pre initialized and ready to respond immediately.
aws lambda put-provisioned-concurrency-config --function-name my-function --qualifier prod --provisioned-concurrent-executions 5
Real-world example A customer facing API backed by Lambda enables Provisioned Concurrency to keep several execution environments warm at all times, ensuring consistent low latency responses for users even during periods of low traffic when a cold start would otherwise be more likely.

Common follow-ups: How much additional latency does a typical cold start actually add?;What is the cost tradeoff of enabling Provisioned Concurrency compared to accepting occasional cold starts?

Auto Scaling Groups;Amazon API Gateway
All 7 Lambda & Serverless questions