vSAN ESA Performance Degradation Due to Incompatible Non-NVMe or SAS Storage Devices

Products

• VMware vSAN

Issue/Introduction

VMware vSAN Express Storage Architecture (ESA) exhibits severe performance degradation, high latency, and stability issues. You observe the following symptoms in your environment:

• vSAN ESA storage pools configured with SAS/SATA SSDs or HDDs incorrectly marked as SSD devices.
• vmkwarning logs indicate significant I/O latency increases:

2026-03-25T08:59:16.536Z Wa(180) vmkwarning: cpu27:2098036)WARNING: ScsiDeviceIO: 1780: Device <REDACTED_DEVICE_ID> performance has deteriorated. I/O latency increased from average value of 1767 microseconds to 57341 microseconds.

• The vdq -i -H command lists non-NVMe devices within the SingleTierDisks group.

Environment

VMware vSAN 8.x ESA

Cause

The storage pool uses non-NVMe devices (such as TOSHIBA KPM5XVUG480G SAS SSDs) or HDDs forced as SSDs,
which do not meet the strict NVMe-only requirement for vSAN ESA.

Resolution

1. Verify Device Compliance: You must use the following commands to confirm if the devices in your storage pool are compliant:
• Run localcli vsan storagepool list to identify devices participating in the singleTier pool.
• Run localcli storage core device list -d <REDACTED_DEVICE_ID> to check the vendor and model of the disks.
2. Identify Non-Compliant Hardware: You check the identified models against the VMware Compatibility Guide (VCG). vSAN ESA strictly requires high-performance NVMe SSDs.
3. Remediate the Configuration:
• You evacuate all data from the impacted host using the "Full Data Migration" mode to ensure data availability.
• You remove the non-compliant SAS/SATA or HDD devices from the vSAN storage pool.
• You replace the incompatible hardware with Broadcom-approved NVMe SSDs listed on the vSAN ESA VCG.
• You re-add the new NVMe devices to the vSAN ESA storage pool to restore performance.

Additional Information

• vSAN ESA Hardware Requirements