Cheatsheet

AWS Serverless

Lambda, DynamoDB and the pieces around them from the CLI: create and call functions, read logs, model and query a table.

Practice this path

Lambda: create and call

```
zip function.zip index.js
```
Package the code (Lambda takes a zip)

aws lambda create-function --function-name hello --runtime nodejs20.x --handler index.handler --role ROLE_ARN --zip-file fileb://function.zip

Create a function

aws lambda invoke --function-name hello --payload '{"name":"Ann"}' --cli-binary-format raw-in-base64-out out.json

Call it with an event and save the reply

aws lambda update-function-code --function-name hello --zip-file fileb://function.zip

Ship a new version of the code

aws lambda update-function-configuration --function-name hello --timeout 15 --memory-size 256

Change timeout and memory (CPU grows with memory)

aws lambda list-functions --query "Functions[].FunctionName"

List functions

aws lambda delete-function --function-name hello

Remove a function

Lambda: handler shape

exports.handler = async (event) => ({ statusCode: 200, body: JSON.stringify({ ok: true }) });

A minimal Node.js handler that returns an HTTP-style reply

def handler(event, context):
    return {"statusCode": 200, "body": "ok"}

The same in Python

```
process.env.TABLE_NAME
```
Read configuration from an environment variable (set with update-function-configuration --environment)
```
context.getRemainingTimeInMillis()
```
How long is left before the timeout (Node.js)

Logs

aws logs tail /aws/lambda/hello --follow

Watch a function's log as it runs

aws logs tail /aws/lambda/hello --since 1h --filter-pattern ERROR

Only the errors from the last hour

aws logs describe-log-groups --log-group-name-prefix /aws/lambda/

List the log groups of your functions

console.log(JSON.stringify({ level: "info", msg: "saved", id }))

Log JSON so it can be filtered and counted

DynamoDB: tables

aws dynamodb create-table --table-name Users --attribute-definitions AttributeName=id,AttributeType=S --key-schema AttributeName=id,KeyType=HASH --billing-mode PAY_PER_REQUEST

Create an on-demand table with a partition key

aws dynamodb describe-table --table-name Users

Key schema, status and size

```
aws dynamodb list-tables
```
List tables

aws dynamodb delete-table --table-name Users

Delete a table

DynamoDB: items

Values carry a type: S string, N number, BOOL, L list, M map.

aws dynamodb put-item --table-name Users --item '{"id":{"S":"u1"},"name":{"S":"Ann"}}'

Write (or replace) an item

aws dynamodb get-item --table-name Users --key '{"id":{"S":"u1"}}'

Read one item by key (fast and cheap)

aws dynamodb update-item --table-name Users --key '{"id":{"S":"u1"}}' --update-expression "SET age = :a" --expression-attribute-values '{":a":{"N":"31"}}'

Change attributes without replacing the item

aws dynamodb query --table-name Orders --key-condition-expression "pk = :p" --expression-attribute-values '{":p":{"S":"u1"}}'

Read many items that share a partition key

```
aws dynamodb scan --table-name Users --max-items 20
```
Read the whole table (slow and costly: avoid on big tables)

aws dynamodb delete-item --table-name Users --key '{"id":{"S":"u1"}}'

Delete an item

```
--condition-expression "attribute_not_exists(id)"
```
Add to put-item: only create if it does not exist yet

Design reminders

```
Design tables around your queries
```
Choose keys for how you read the data; there are no joins
```
Least privilege for the function role
```
Give each Lambda only the actions and tables it needs
```
Idempotent handlers
```
Events can be delivered twice: make a repeat harmless
```
Cold start
```
The first call after idle is slower; keep packages small

Cheatsheet for the AWS Serverless path · Deboistech · All cheatsheets