Skip to content

Serverless Cube.js backend infrastructure on AWS

License

Notifications You must be signed in to change notification settings

jimjoes/serverless-cubejs

 
 

Repository files navigation

serverless-cubejs

Serverless Cube.js backend infrastructure on AWS.

Architecture

This project uses

  • API Gateway HTTP APIs for hosting the Cube.js backend API
  • Athena for querying data in S3
  • Aurora Serverless for pre-aggregations
  • ElastiCache for query result caching

All of that runs in a preconfigured VPC. You can also have a look at the stack diagram in docs/stack.drawio.

AWS pricing applies, please make yourself familiar with the pricing of the deployed resources BEFORE you deploy the stack. This will not be covered by the AWS Free Tier!

Usage

Prerequisites

You'll need an AWS account (obviously), and a readily installed Serverless framework, along with local AWS credentials.

Cube configuration

You need to adapt (or add) the schemas for your cube in the schema folder appropriately (see Cube.js docs).

Deployment

$ sls deploy --aurora-password YOUR_AURORA_PASSWORD --s3-data-bucket YOUR_S3_BUCKET_NAME --cube-secret YOUR_CUBE_SECRET

The default stage that will be used is dev. You can also specify if via the --stage commandline option.

Removal

$ sls remove

It's possible that you have to manually clean some resources, e.g. the S3 bucket with the Athena query results.

About

Serverless Cube.js backend infrastructure on AWS

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • JavaScript 100.0%