HAR in AWS Lambda Deployments
X
Xylophone gold Bobolink
Current Behavior:
Harness CD’s Lambda Swimlane currently requires artifacts to be hosted in Amazon S3. When using a custom artifact source, the deployment process expects the script to return:
Artifact version
S3 bucket name
S3 file path
This setup works technically, but when the source of truth is Harness Artifact Registry (HAR), customers are forced to:
- Use HAR's Versions API to retrieve artifact metadata.
- Manually copy/download the artifact from HAR to S3.
3.Return the S3 location via the custom script for Harness CD to consume.
This results in duplicate storage, increased complexity, and a suboptimal experience.
Proposed Enhancement:
Introduce first-class support for HAR in AWS Lambda deployments. This would allow Harness CD to:
Directly fetch the artifact from HAR.
Eliminate the need to stage artifacts in S3.
Simplify the deployment workflow for users relying on HAR.
Value/Impact:
Reduces operational overhead for users.
Avoids data duplication between HAR and S3.
Enhances the value proposition of HAR as a unified artifact source across deployment types.
Log In