Block Storage charges for provisioned storage capacity regardless of whether the disk is attached to an ECS instance or actively used. To reduce costs, regularly audit your cloud disks and snapshots, then take one of three actions: delete what you no longer need, right-size what you do, or automate snapshot retention.
How it works
When an ECS instance is stopped or terminated, attached cloud disks are not automatically deleted. Those disks remain provisioned and continue to incur charges until you explicitly delete them. The same applies to disks you detach manually.
Cost optimization therefore requires active management. The three sections below correspond to three distinct situations:
| Situation | Action |
|---|---|
| Disks that are unattached or underused | Delete them |
| Disks that are overutilized, or Enhanced SSDs (ESSDs) whose performance level can be reduced | Adjust capacity or performance level |
| Snapshots accumulating beyond your retention needs | Set a retention period to auto-delete old ones |
Delete unattached or unused cloud disks
The fastest way to reduce Block Storage costs is to find and delete cloud disks that are no longer attached to any ECS instance.
To identify candidates for deletion:
Unattached disks: Check for disks with no associated ECS instance. These incur charges without serving any workload.
Inactive disks: Check whether read and write operations have been performed on a disk in the past few weeks. A disk with no recent I/O is a strong signal it is no longer needed.
For non-production environments, delete a disk if it has had no read or write activity for several weeks, or has been unattached for one month or more.
Adjust cloud disk capacity
For cloud disks that are overutilized — meaning they are running out of space — scale up the disk online or offline to increase capacity.
For ESSDs, the performance level (PL) provides additional flexibility:
Upgrade PL online: Increase the PL to meet higher performance or capacity requirements without stopping the instance.
Downgrade PL online: For pay-as-you-go ESSDs, reduce the PL to lower storage capacity and cost. This is useful for non-production environments where peak performance is not required.
To restore a cloud disk to its initial state, re-initialize the disk. This resets the disk to the state it was in when created.
Manage snapshot retention
Automatic snapshot policies that run on a daily or weekly schedule accumulate snapshots over time. Without a retention limit, old snapshots continue to use storage and incur charges.
Set a retention period in your automatic snapshot policy. When a snapshot exceeds the retention period, it is automatically deleted. Deleting snapshots does not affect Block Storage.