Managing Oversized Ebs Volumes in Production

Oversized EBS volumes are common in production; here's how to diagnose, remediate, and prevent them effectively.

May 30, 2026 JR

1 minute read

Oversized EBS volumes are common in production; here’s how to diagnose, remediate, and prevent them effectively.

Start by identifying volumes where allocated space far exceeds actual usage. Use these methods:

Check current usage:

df -h /mount/path  # For mounted volumes

List EBS volumes:

aws ec2 describe-volumes --filters 'Name=attachment.instance-id,Values=i-1234567890'

Monitor via CloudWatch:
Use VolumeIdleWrite and VolumeIdleRead metrics to spot underutilized volumes.

Create a snapshot of the oversized volume.
Create a new volume from the snapshot with a smaller size (e.g., 100GiB → 50GiB).

Attach the new volume to the same instance:

aws ec2 attach-volume --volume-id vol-1234567890abcdef0 --instance-id i-1234567890abcdef0 --device /dev/xvdf

Transfer data from old to new volume:

rsync -avx /mnt/old_volume/ /mnt/new_volume/  
# OR use dd for block-level copy (slower, more thorough):  
dd if=/dev/xvdf of=/dev/xvdg bs=1M

Detach old volume and reattach new volume with the original device name.
Expand the filesystem (if needed):
```
resize2fs /dev/xvdf  # For ext4  
```

blog