Error: Virtual machine has flash read cache reservation during vROps optimization

Products

• VMware vSAN

Issue/Introduction

A VMware vSAN environment reports optimization failures when running reports in vRealize Operations (vROps). The report incorrectly flags virtual machines with the error "Virtual machine has vSAN flash read cache reservation," blocking automated cluster balancing. This occurs despite manual vMotion working correctly and VM Storage Policies (SPBM) being set to 0% cache reservation.

Environment

• vRealize Operations (vROps) 8.18
• VMware vSAN 8.0 Update 3 (ESA or All-Flash)
• VMware vCenter Server 8.0.x

Cause

The issue is caused by stale metadata or legacy property flags retained in the vCenter inventory or the vROps Postgres database. When a cluster is migrated from Hybrid to All-Flash storage, the cacheReservation metric may remain in the management layer despite the physical vSAN objects having a 0 reservation.

Resolution

1. Verify Physical Object State: Log in to an ESXi host and run the following to confirm physical objects have 0 reservation:

   localcli vsan debug object list | grep -A1 "cacheReservation"

2. Force Inventory Metadata Refresh:
   • Navigate to Cluster > Monitor > vSAN > Virtual Objects.
   • Select affected VMs and click Check Compliance.
   • If non-compliant, click Reapply.
3. vROps Database Cleanup:
   • If the error persists in vROps reports, engage the vROps support team.
   • Perform a forced inventory discovery or refresh the vROps Postgres database metrics.