Benchmark: EBS Checks
Thrifty EBS Benchmark
Thrifty developers keep a careful eye for unused and under-utilized EBS volumes. Elastic block store is a key component of hidden cost on AWS, and this benchmark looks for EBS volumes that are unused, under-utilized, out-dates and oversized.
Variables
| Variable | Description | Default |
|---|---|---|
| ebs_snapshot_age_max_days | The maximum number of days snapshots can be retained. | 90 days |
| ebs_volume_avg_read_write_ops_high | The number of average read/write ops required for volumes to be considered frequently used. This value should be higher than ebs_volume_avg_read_write_ops_low. | 500 ops/min |
| ebs_volume_avg_read_write_ops_low | The number of average read/write ops required for volumes to be considered infrequently used. This value should be lower than ebs_volume_avg_read_write_ops_high. | 100 ops/min |
| ebs_volume_max_iops | The maximum IOPS allowed for volumes. | 32,000 IOPS |
| ebs_volume_max_size_gb | The maximum size (GB) allowed for volumes. | 100 GB |
Usage
Install the mod:
mkdir dashboardscd dashboardspowerpipe mod initpowerpipe mod install github.com/turbot/steampipe-mod-aws-thriftyStart the Powerpipe server:
steampipe service startpowerpipe serverOpen http://localhost:9033 in your browser and select EBS Checks.
Run this benchmark in your terminal:
powerpipe benchmark run aws_thrifty.benchmark.ebsSnapshot and share results via Turbot Pipes:
powerpipe benchmark run aws_thrifty.benchmark.ebs --shareControls
- Old EBS snapshots should be deleted if not required
- EBS volumes attached to stopped instances should be reviewed
- Are there any EBS volumes with low usage?
- Still using gp2 EBS volumes? Should use gp3 instead.
- EBS volumes with high IOPS should be resized if too large
- Still using io1 EBS volumes? Should use io2 instead.
- EBS volumes should be resized if too large
- What provisioned IOPS volumes would be better as GP3?
- Are there any unattached EBS volumes?