Deploy serverless functions to aws lambda
Problem to solve
We offer the ability for users to easily deploy serverless functions to Knative, however, we don't offer the same level of convenience for lambda (the most popular serverless solution today). Lambda users have to configure custom
gitlab-ci.yml, IAM, S3, etc.
Sasha, Software Developer, https://design.gitlab.com/research/personas#persona-sasha
Devon, DevOps Engineer, https://design.gitlab.com/research/personas#persona-devon
Example user flow:
- User enters their AWS credentials as CI variables
gitlab-serverless.ymlfile defining lambda functions (with some lambda identifier)
- GitLab recognizes the lambda identifier and
gitlabktlpackages into necessary format (ie
Create a lambda-specific CI flow in
gitlabktl(similar to Knative) that is triggered when user has
gitlab-serverless.yml. See flow above.
Provide link to cloudwatch logs within serverless page.
- Requests will need to be proxied through backend for auth reasons and maybe even for CORS reasons
What does success look like, and how can we measure that?
Create a mechanism to identify which
provider is entered on
serverless.yml. Currently, we identify serverless usage by the deployment of Knative to a cluster, this would be an opportunity to refine to parse the
provider defined in the