This little tool helps to estimate the amount of AWS logs your AWS account is generating. Currently it supports:

• VPC Flow Logs (via CloudWatch Logs or S3)
• CloudTrail

It tries to guestimate the VPCFL traffic using 2 heuristics:

1. Calculate it from CloudWatch Logs (using the new CWL Insights query lang) - OR-
2. Estimate it from (zipped) S3 daily sample (using an 8X COMPRESSION_RATIO constant)

The tool is written in python3 and requires a working python3 environment. To verify:

python3 --version

If your default python installation is python3 then you can replace python3 with just python

The tool assumes that a working AWS cli / profile was already set. If in doubt - please consult boto guide : https://boto3.amazonaws.com/v1/documentation/api/latest/guide/configuration.html
Your AWS user / role should have some readonly permissions to perfrom a few EC2 describe operations, CloudWatchLogs Insight queries, and to read relevant S3 buckets. This is the list of IAM permissions that is required. It can be added to the relevant user / role as an independant policy.

{
    "Version": "2012-10-17",
    "Statement": [
        {
            "Sid": "storagecalculator",
            "Effect": "Allow",
            "Action": [
                "ec2:DescribeVpcs",
                "ec2:DescribeRegions",
                "ec2:DescribeFlowLogs",
                "s3:ListBucket",
                "s3:GetObject",
                "logs:StartQuery",
                "logs:GetQueryResults"
            ],
            "Resource": "*"
        }
    ]
}

the tool assumes the standard AWS (boto) SDK setup. All authentication setup should be done externally using environment variables. If you wish to use the default AWS profile (or your AWS instance's role):

python3 calculator.py

If you wish to use a different profile, use the relevant environment variable as follows:

export AWS_PROFILE=<MY PROFILE_NAME> && python3 calculator.py

Similarly, it is possible to set the AWS_ACCESS_KEY_ID (and AWS_SECRET_ACCESS_KEY) variables, but it is left out of this readme.

Pretty self explanatory. The scripts enumerates relevant CWL groups and relevant S3 buckets and estimates the daily/monthy data volume (non compressed).