This article describes the process for transitioning a VMware vSAN cluster running Original Storage Architecture (OSA) to Express Storage Architecture (ESA).
Direct or in-place upgrade/migration from vSAN OSA to vSAN ESA is not supported because OSA and ESA utilize fundamentally different data paths, storage mechanisms, and on-disk formats. OSA uses disk groups composed of dedicated cache and capacity devices, whereas ESA utilizes a unified storage pool architecture with high-performance NVMe devices.
Direct conversion or in-place upgrade from vSAN OSA to vSAN ESA is not natively supported. Transitioning to ESA requires constructing a new ESA-enabled cluster or performing a destructive re-initialization of host storage devices.
Since an in-place migration is not possible, choose one of the following two supported transition methods based on available hardware resources.
Method 1: Side-by-Side Migration (New Hardware Available)
This method is recommended when new ESA-compatible hardware is available to build a parallel cluster.
Method 2: Host Reuse & Destructive Re-initialization (Re-using Existing Hosts)
If existing hosts are being repurposed for the vSAN ESA cluster, all storage devices must be cleared of previous partition metadata and placed into a Clean or Unmounted state before ESA can claim them into the Storage Pool.
Note: Re-using hosts requires wiping all existing disk groups and partitions. Ensure all Virtual Machines are evacuated off the host prior to deleting disk groups. Any remaining data on the host's storage devices will be permanently lost.