VPC flow logs do aggregation already, but across larger time windows you have to do it yourself. This PR adds a library function and CLI action to compute time windows and traffic totals for all records from a query.
This is unidirectional aggregation and thus keeps the same field names as are in the original flow records. I may do bidirectional aggregation in another PR later.
Coverage remained the same at 100.0% when pulling 63977dc787f289f8f62582a079efc67701cd9c8b on aggregated_records into 18eafa0eb3dda9e68c037a88b0f100c2fc723329 on master.
VPC flow logs do aggregation already, but across larger time windows you have to do it yourself. This PR adds a library function and CLI action to compute time windows and traffic totals for all records from a query.
This is unidirectional aggregation and thus keeps the same field names as are in the original flow records. I may do bidirectional aggregation in another PR later.