This project is to resolve the following questions/tasks:
[ ] Restore a Table from S3 in a parquet data format
[/] Table DynoamoDB Backups, can these be automagicaly streamed to other accounts within an aws org without undue pain? {AWS Backup Service ❤️}
[/] Can a table that is defined within a CloudFormation be restored and ingested into a happy CloudFormation Stack? If so What changes are need to allow us to do this?
[ ] Can the default aws SSO admin role have the delete table action permission removed easily without causing hassle?
These questions/tasks need to be resolved before giving the go ahead for a pile of effort to be put in loading up production with sensitive data for meetBel. We have a impending release deadline an these questions are currently a blocker.
Cake HTTP API allows the caller to get cakes, cakes are stored in DynamoDB under the Cake partition key, simple JS lambda to scoop this out. This is
an absolute minimal application that I am using to dry-run restoring a DynamoDB table for a running application.
So far I have switched out the table to a recovered version of the table and imported that new DynamoDB table into this stack and maintained a in-sync
DynamoDB stack state.
I would like to at some point restore the state of the dynamoDB from a parquet dump in S3, we already have a data pipeline that persists the table in the Parquet format for our "data science" requirements, so if this can also be used for restore purposes that would be great.
I have also set up the aws Backup Service with this table and restored from a back up in the vault.
This does resolve the questions I had before giving the go ahead for production data to be populated. 🤔
Build and lint (in service directory):
npm run build
Package:
aws cloudformation package --s3-bucket your-cf-bucket --output-template-file packaged.yaml
Deploy:
aws cloudformation --template-file /FULL/PATH/TO/packaged.yaml --stack-name YOUR_STACK_NAME --region eu-west-1 --capabilities CAPABILITY_IAM