Storage vMotion limitation for boot disks of Clustered Virtual Machines

Products

• VMware vCenter Server
• VMware vSphere ESXi

Issue/Introduction

* Storage vMotion operations on cluster VM boot disks trigger cluster failover events

Environment

VMware vSphere versions 4.0 and newer

Cause

The fundamental limitation stems from how Storage vMotion interact with clustered VMs:

1. VM Quiescence Requirement
   • All vMotion operations require brief VM quiescence during the final switchover
   • For Storage vMotion, this quiescence period is longer due to storage data transfer
   • Even non-RDM disks (like boot volumes) require complete OS and application quiescence
     
     
2. Cluster Sensitivity
   • Cluster software monitors inter-node communication latency
   • Any quiescence exceeding cluster timeout thresholds triggers failover
   • This behavior is by design to ensure high availability
     
     
3. RDM Considerations
   • RDM disks maintain SCSI-3 persistent reservations for cluster coordination
   • Storage operations can disrupt these reservations
   • Disk device reconnection during vMotion can trigger cluster protective measures

Resolution

For Non-RDM Boot Disks in Cluster VMs, Storage vMotion of boot disks will trigger cluster failovers. Plan maintenance windows accordingly.

Additional Information

Best Practices

• Always maintain current backups before migration
• Document all RDM mappings thoroughly
• Test procedure in non-production environment first
• Schedule adequate maintenance window
• Have rollback plan ready

Related Documents

• Migrating virtual machines with Raw Device Mappings (RDMs)