Most Amazon Web Services pain is self-inflicted and predictable. Here are the mistakes we see most, and the fix for each.
Mistake 1: Long-lived access keys instead of SSO/OIDC roles
Fix: Use SSO for humans and OIDC roles for CI - no static keys. Make it the default in your templates and enforce it in review or CI so the mistake cannot come back.
Mistake 2: Over-broad IAM policies with wildcard actions and resources
Fix: Write least-privilege IAM and manage everything via IaC. Make it the default in your templates and enforce it in review or CI so the mistake cannot come back.
Mistake 3: Public S3 buckets and security groups open to 0.0.0.0/0
Fix: Turn on CloudTrail, GuardDuty and budget alerts. Make it the default in your templates and enforce it in review or CI so the mistake cannot come back.
A safer baseline
A least-privilege S3 read policy as Terraform.
data "aws_iam_policy_document" "read_assets" {
statement {
actions = ["s3:GetObject"]
resources = ["arn:aws:s3:::acme-assets/*"]
}
}
resource "aws_iam_policy" "read_assets" {
name = "read-assets"
policy = data.aws_iam_policy_document.read_assets.json
}Want senior help putting Amazon Web Services to work - or a team bootcamp on it? CodeWizard runs Israel's largest DevOps community and embeds named senior engineers with a transparent hours log. Book a free consult or explore the Academy.