This PR fix the issues,
- which related to AWS api limits(https://docs.aws.amazon.com/AmazonCloudWatch/latest/logs/cloudwatch_limits_cwl.html). Basically, if you set granularity in rds console for small time interval, like 5/10 seconds, and have 2 mins, you will faced with
time="2018-08-01T08:12:26Z" level=error msg="unable to get logs for instance : ThrottlingException: Rate exceeded\n\tstatus code: 400, request id: a02da727-9562-11e8-86c9-0d07b3aaf681" source="enhanced.go:109"
And you will have spots in promethues graph. it will be good to have this interval set via configuration file.
- In basic metrics, it will be nice to see cpu utilization set to total or user instead of idle.