An AWS Professional Service open source initiative | aws-proserve-opensource@amazon.com
AWS CodeSeeder
The aws-codeseeder project enables builders to easily “seed” python code to AWS CodeBuild for execution in their cloud environments.
The library and its CLI utility can be used to simplify development and deployment of complex Infrastructure as Code projects. These projects usually have many dependencies and require mutiple CLI utilities to orchestrate their deployments. For example, a project that deploys networking resources with the AWS CDK, an Amazon EKS Cluster with the eskctl CLI utility, and Kubernetes applications with kubectl and helm CLI utilities.
The aws-codeseeder elminates the need to install and configure libraries and utilities locally or on a build system (i.e. Jenkins). Instead the library enables builders to easily execute an AWS CodeBuild instance with the utilities they require then seed local python code to, and execute it within, the CodeBuild instance. By bundling and executing local python code in AWS CodeBuild, aws-codeseeder can enable GitOps type deployments of complex, mixed technology projects.
How It Works
Each CodeSeeder client project does a one-time deployment of a named Seedkit of AWS resources. Deployment can be done with the CodeSeeder module or the CLI utility, resources include:
AWS CodeBuild Project: remote executor
S3 Bucket: local resources are bundled and uploaded, acts as Source to the CodeBuild executions
KMS Key: for optional encryption of bundles and source
AWS CodeArtifact Domain and Repository: for optional serving of shared Python packages
IAM Role: default Role used by the CodeBuild executions
IAM Managed Policy: grants access to Seedkit resources
In the CodeSeeder client project, builders use the codeseeder.remote_function decorator to mark code for execution by the remote CodeBuild Project. When functions marked with the remote_function decorator are called locally, the CodeSeeder library intercepts the call, bundles the local project and dependencies, pushes the bundle to S3, starts a CodeBuild execution using the bundle in S3 as Source, then executes the marked function remotely in the CodeBuild project.
