We have been testing Functionbeat within our environment and the most frustrating part of the deployment is not being able to find detailed documentation on how to perform this cleanly. The approach seems to be using the manager or using the manager to export a package.zip that seems tightly coupled with the manager configuration file... we need the granularity to conform with naming conventions and build standards.
We have reviewed https://github.com/elastic/beats/issues/13372 which provides some sort of guidance but it does not feel sufficient...
- The functionbeat.yml file that zipped with functionbeat-aws binary, which fields are optional and mandatory?
- What is the bare minimum required by the binary to execute the Lambda function?
- As triggers, VPC configuration, IAM role are defined on the Lambda function itself, can this information (including path, keystore, deploy_bucket, endpoints etc.) be removed in the YAML file?
My main concern is that if I'm deploying Functionbeat with a custom IAM role, security group, VPC configuration, how do I pass all of these parameters into the configuration file without creating a separate pipeline or hardcoding values into the functionbeat.yml file?