VM replication is stuck during the initial synchronization

Products

• VMware vSphere ESXi

Issue/Introduction

Symptoms:

• VM replication using vSphere Replication may become stuck during the initial synchronization.

 

 

Environment

vSphere Replication 9.x

Cause

This issue is caused by checksum and read errors on the VMDK that prevent vSphere Replication (VR) from accessing the required disk blocks, and their occurrence across multiple datastores indicates underlying corruption or unreadable sectors within the VMDK.

• The ESXi host’s vmkernel logs consistently report failures while attempting to read checksum data from the affected VMDK:

2025-11-12T07:21:49.143Z Wa (180) vmkwarning: cpu3:2098075) WARNING: Hbr: 1528: Failed to read checksum data (DiskID=RDID-########-175d-####-853b-###########
(offset15100870656) (extentSize=8192) (count=8): I/0 error

Resolution

Perform a VMDK integrity check and repair, and if corruption is confirmed, restore the VM from a known good backup or snapshot before reconfiguring replication. If no valid backup is available, consider cloning the VM or creating a new VMDK and migrating the data at the OS or application level to ensure it resides on a healthy disk.