Resource Catalog
cloudrift analyzes 44 resource types divided into two categories:
waste— money being spent now, eliminable by removing/detaching the resource. Contributes to the headline total and the CI gate.optimization— a savings opportunity that keeps the resource, shown separately and never gated.
For the 9 most recent “vertical” scanners (SQS DLQ, Aurora Serverless, SageMaker, EKS, Dev/PR ghost environments) with config/risk deep dives, see Vertical Scanners.
Core resources (waste)
| Resource | Waste condition | Estimated cost (us-east-1) |
|---|---|---|
| EBS Volumes | Unattached (state: available) |
gp3: $0.08/GB-mo · gp2: $0.10/GB-mo · io1: $0.125/GB-mo |
| Elastic IPs | Unassociated (no EC2/NAT binding) | $3.60/month fixed |
| RDS Instances | Stopped (storage still billed) | gp2/gp3: $0.115/GB-month |
| Load Balancers | No registered targets (ALB/NLB) | ~$16.20/month fixed |
| EC2 Instances | Stopped (attached EBS still billed) | Sum of attached EBS volumes |
| EBS Snapshots | Source volume deleted (orphans) | $0.05/GB-month |
| NAT Gateways | Zero outbound traffic in the last 14 days | ~$32.40/month fixed |
| EBS Volumes (idle) | Attached but zero I/O in the last 14 days | Same as above by type |
| CloudWatch Log Groups | No retention policy configured | $0.03/GB-month |
| Orphaned ENIs | Status: available (not attached) |
$0 (hygiene flag) |
| EFS File Systems (unused) | No mount targets or zero I/O in 14 days | $0.30/GB-month |
| ElastiCache Clusters (idle) | Zero connections in 14 days | Full node-hour cost |
| Redshift Clusters (idle) | Zero DB connections in 14 days | Node-hour × nodes |
| OpenSearch Domains (idle) | Zero requests in 14 days | Instance-hour × instances |
| MSK Clusters (idle) | Zero broker traffic in 14 days | Broker-hour × brokers |
| FSx File Systems (idle) | Zero I/O in 14 days | $0.093–$0.14/GB-month |
| DocumentDB (idle) | Zero connections in 14 days | Full instance-hour cost |
| Neptune (idle) | Zero query traffic in 14 days | Full instance-hour cost |
| Amazon MQ (idle) | Zero network traffic in 14 days | Broker-hour (×2 for Multi-AZ) |
| WorkSpaces (idle) | AlwaysOn, no user connection in 30 days | Full bundle monthly cost |
| VPN Site-to-Site (idle) | Zero tunnel traffic in 14 days | ~$36.50/month |
| Transit Gateway (idle) | Zero traffic in 14 days | ~$36.50/month |
| Kinesis Streams (idle) | Provisioned, zero records in 14 days | ~$10.95/month per shard |
| SQS DLQ (abandoned) | Oldest message > 14 days | $0 (hygiene flag) |
| CloudWatch Log Groups (orphaned Lambda) | Lambda function no longer exists | $0.03/GB-month |
| SageMaker Notebook Instances (idle) | InService, max CPU ≤ 2% over 7 days |
Full instance-hour cost |
| SageMaker Endpoints (idle) | InService, zero invocations over 7 days |
Full instance-hour cost × instance count |
| AMI (unused) | Owned AMI not referenced by any instance or launch template | Cost of underlying EBS snapshots, $0.05/GB-month |
| ECR Images (untagged) | Dangling image (no tags) in a repository | $0.10/GB-month |
| S3 Multipart uploads (abandoned) | Incomplete multipart upload, never completed or aborted | $0.023/GB-month |
| RDS Manual Snapshots (old) | Manual snapshot older than the grace period | $0.095/GB-month |
| Secrets Manager Secrets (unused) | Never accessed, or not accessed in the last 30 days | $0.40/secret/month fixed |
| CodePipeline Pipelines (stale) | No execution within the grace period (or never executed) | $1.00/month fixed per pipeline |
Optimizations
| Resource | Condition | Estimated saving |
|---|---|---|
| EBS gp2→gp3 | In-use gp2 volume upgradeable | ≈ $0.02/GB-mo |
| EC2 (underutilized) | Running, max CPU ≤ 5% over 14 days | Real price difference to one size down in the same family, $0 if not derivable |
| RDS (underutilized) | Max CPU ≤ 5% over 14 days | Real price difference to one size down in the same family, $0 if not derivable |
| S3 Buckets (no lifecycle) | No lifecycle configuration | $0 (hygiene flag, no real basis for a figure) |
| Lambda (underutilized) | Zero invocations in 7 days | $0 (hygiene, pay-per-use) |
| DynamoDB (overprovisioned) | Read/write utilization < 10% in 7 days | Real price difference between current and recommended capacity |
| Aurora Serverless v2 | Min ACU overprovisioned | (Min ACU − suggested) × $87.60/ACU-month |
| SageMaker Models (orphaned) | Not referenced by any endpoint | $0 (no real basis for a figure) |
| EKS Node Groups | CPU requested < 30% of allocatable | (nodes − suggested) × instance price |
| EKS Orphaned PVC | Kubernetes-provisioned EBS unattached | EBS price by type |
| Dev/PR Environments (ghost) | All resources inactive for 7+ days | Estimated total of the group |
See also
- Quick Start — installation and IAM permissions
- Vertical Scanners — deep dives on the 9 most recent scanners
- Pricing — how prices are resolved and the
confidenceaxis