Provisioned Space is higher than expected for non-persistent VDI VMs.
search cancel

Provisioned Space is higher than expected for non-persistent VDI VMs.

book

Article ID: 435010

calendar_today

Updated On:

Products

VMware vSAN

Issue/Introduction

  • You have a VDI VM reporting Provisioned Space (Datastores Inventory View -> VMs Tab) vastly higher (2 -3 times) than the configured size of the virtual disks, as shown in: Right-click VM -> Edit Settings).
  • The provisioned space may be lower while the VM is powered off.
  • There are no snapshots showing in the Snapshot Manager for the VMs (Right-click VM -> Snapshots -> Snapshot Manager -> Manage Snapshots).

 

Environment

  • VMware vSphere 8.x

Cause

  • This behavior is expected due to the multi-layered architecture used by VDI solutions (such as Citrix MCS) to manage non-persistent clones. The "Provisioned Space" reflects the total potential footprint of all underlying components required to run the VM, including:

    • Base Replicas: A full copy of the golden image disk.

    • Snapshots/REDO Logs: Non-persistent VMs utilize hidden snapshots and REDO files (write caches) for operation. Each of these counts the full "configured" size of the base disk toward the provisioned total, even if thin-provisioned.

    • vSAN Overhead: In vSAN environments, policy-based protection (RAID) and architecture (ESA) multipliers significantly increase the calculated provisioned space.

Resolution

  • To verify if the reported space is accurate, calculate the expected footprint based on your environment's architecture and storage policies.

 

 

1. Identify the Multipliers 

Use the following table to determine the overhead based on your specific vSAN configuration:

ComponentMultiplier / Overhead
vSAN ESA Architecture1.131 (~13.1% for performance leg)
RAID-1 (Mirroring)2.0 x
RAID-5 (Erasure Coding)1.33x
RAID-6 (Erasure Coding)1.5x

 

 

 

2. Example Calculation: 

For a VM with 16 GB RAM (no reservation) and a single 100 GB virtual disk using RAID-6 on vSAN ESA, the math is as follows:

      1. VDI Layers: (Base replica + Snapshot + REDO log) = 300 GB

      2. Apply ESA Overhead: 300 GB x 1.131 = 339.3 GB

      3. Apply RAID-6 Policy: 339.3 GB} x 1.5 = 508.95 GB

      4. Add Swap File: 508.95 GB + 16 GB swap = ~524.95 GB Total Provisioned

 

[!IMPORTANT]

These snapshots and REDO logs are hidden from the Snapshot Manager. They are managed automatically by the VDI solution and should not be removed manually via the CLI, as this can cause data corruption or pool failure.

 

Additional Information