Excessive storage consumption on target datastore during HCX Bulk Migration

Products

• VMware HCX

Issue/Introduction

During an active HCX Bulk Migration, the target datastore may show significantly higher storage consumption than the provisioned size of the source Virtual Machine (VM).

Environment

VMware HCX 9.x
VMware vSphere 7.x/8.x

Cause

Target datastore space decreases continuously during the replication phase. The space consumed on the destination exceeds the 1.3x - 1.5x best practice recommendation.
Large hbrdisk.RDID.snapshot.vmdk files are present in the target VM directory. 

HCX Bulk Migration utilizes vSphere Replication (vSR) which creates redo logs to capture data churn on the source VM during the 2-hour RPO cycles. For high-churn workloads or migrations spanning several days, these redo logs accumulate on the target datastore. These files are not consolidated until the switchover phase is initiated.

 

Resolution

Verify the data churn rate on the source VM using HCX performance metrics. Confirm the presence of multiple hbrdisk.RDID.snapshot.vmdk files on the target datastore. Schedule and execute the switchover for the impacted VM. Monitor the "Offline Sync" and "Instantiation" tasks during switchover, as this is when the redo logs are merged into the base disk and space is reclaimed. If storage is critically low, consider increasing the datastore capacity or moving other workloads to free up space until the migration completes.