aws-samples

AWS Cheatsheet

//cli

//cognito

--create a user
aws cognito-idp sign-up \
  --region REGION \
  --client-id APP_CLIENTID \
  --username APPUSERNAME/EMAIL \
  --password PASSWORD

--confirm a user
  aws cognito-idp admin-confirm-sign-up \
  --region REGION \
  --user-pool-id USERPOOLID \
  --username APPUSERNAME/EMAIL \

//upload file to s3
aws s3 cp  my-lambda-package.zip s3://artifacts-for-lambda-ronin-test


//SAM
//--package app
aws cloudformation package --template-file sam-template.yaml --s3-bucket artifacts-for-lambda-ronin-test --output-template-file sam-output-template.yaml
--or same using "sam package..."
//--deploy app (use the resulting output file)
aws cloudformation deploy --template-file sam-output-template.yaml --stack-name my-serverless-app --capabilities CAPABILITY_IAM
--or same using "sam deploy..."
//--delete the stack
aws cloudformation delete-stack --stack-name my-serverless-app
//--validate yaml
sam validate --template sam-template.yaml
//--test api locally (doesn't work for dynamo etc??)
sam local start-api -t sam-template.yaml
//--invoke lambda locally
sam local invoke MyLambdaFunction -t sam-template.yaml --event ./src/testevent.json
//--invoke events locally (see docs and --help)
same local generate-event [service]


//dynamo -

//modelling
--partition key
--sort key
--primary key is unique and is partition key or partition key + sort key if sort key defined

//RCU (Read Capacity Unit) 
4kb/second of data (for strongly consistent)
8kb/second (for eventually consistent)

//WCU (Write Capacity Unit) 
5kb/second of data (double check?)

//Formulas
RCU - [item size]/4kb (rounded up) (for strongly consistent) (i.e. 8kb would be 2 RCUs)
RCU - ([item size]/4kb)/2 (rounded up) (for strongly consistent) (i.e. 8kb would be 1 RCUs)

//Node Api
putItem - inserts or replaces an item
updateItem - for field-level updates
getItem - gets one by primary key
query - queries on partition and/or sort key, allows for additional filters and sorting on sort key, this is querying within partition
scan - queries on all items, most costly because it ignores partitions
deleteItem - delete item