Goals

When creating a log group, allow the user to choose between the Standard log class for frequently accessed, real-time monitoring logs and the Infrequent Access log class for cost-effective storage and ad-hoc analysis

Non-Goals

setting "delivery" log class (very constrained and therefor larger scope)

Motivation

Key Benefits:

• Cost Allocation: group class offers significant cost savings for logs not needed for real-time monitoring.
• Tailored access: group class provides the flexibility to tailor features to specific access patterns. 

Alternatives

No alternatives.   This feature can only be used when the log group is first created.

Acceptance Criteria

• `log_group_class` is newly added field in the upstream vector image
• `groupClass` can be added to the cloudwatch output spec
• two accepted values are correctly validated and conditions set accordingly
• is optional field and defaults to Standard if not included
• vector config is successfully set based on CLF spec
• create unit tests for validation and spec generation
• e2e tests using our mock??
• Update feature documents
• Update RH documentation

Risk and Assumptions

The aws_cloudwatch_logs sink in Vector needs an update in order to support the storage tier configuration for the log groups.

• POC:  https://github.com/cahartma/vector/pull/1/files

Documentation Considerations

Open Questions

Allowed values are "standard" and "infrequent_access", as these are accepted by vector.   Do we also need to include "infrequentAccess" (camelcase) for consistently, primarily since we don't appear to use any snake case in our enums??

Additional Notes