vSAN Capacity view - Usage by Snapshots

Products

• VMware vSAN

Issue/Introduction

This article introduces the 'Capacity - Usage by Snapshots' view and how the snapshot usage is defined. Below is the screenshot example:

Environment

vSAN ESA

Resolution

Q: What does the 'Usage breakdown - Usage by snapshots' report? 

• Total usage:
  Represents the sum of all physical space usage. It combines the running point (live data), snapshots usage, system usage, etc.

• Snapshot usage:
  

Represents the sum of VMDK snapshots and the usage of VM snapshot memory files. When a snapshot is created, it initially doesn't consume additional space. Instead, it references the blocks of the current running point (live data). Over time, as changes are made to the running point, the original data blocks referenced by the snapshot are retained for consistency, consuming snapshot space.

• VMDK snapshots:
  The disk space used to preserve the state of virtual machine disks at the time the snapshot was taken. As changes are made to the virtual machine, these snapshots grow to track the original data blocks.

• Current data:
  Represents the live portion of the data stored on the vSAN datastore, but distinct from snapshots or backup data. Current data = Total usage - Snapshot usage.

Q: What's the space usage impact of creating a snapshot in vSAN ESA cluster?

No immediate change in total space usage
When a snapshot is created, the live data's space usage is essentially "moved" under the snapshot. However, the total space usage remains the same since no new data is written. The snapshot holds a record of the live data state at the time it was taken.

Subsequent Changes to the Running Point
When new data is written to the running point (or live data is modified), new blocks are allocated to store these changes. This increases the total space usage because now both the new data (in the running point) and the retained old data (in the snapshot) consume storage.

Q: Why Total Space Usage Remains Unchanged Upon Snapshot Creation?

• A snapshot doesn't duplicate data; it just records the current state of the live data. Hence, no new blocks are written immediately.
• Total space usage increases only when new blocks are written due to modifications or additions to the running point.

Example Scenario

• Initial State:

  • Running point: 100GB used
  • Snapshots: 0GB used
  • Total space usage: 100GB

• Snapshot Created:

  • Running point: 0GB (logically, data moved under snapshot for accounting purposes)
  • Snapshots: 100GB used
  • Total space usage: 100GB (unchanged)

• New Data Written (20GB):

  • Running point: 20GB used (new data)
  • Snapshots: 100GB used (unchanged, holding the old 100GB state)
  • Total space usage: 120GB (increased by 20GB)