Disaster Recovery Failback's initial replication takes longer than expected.

Products

• VMware HCX

Issue/Introduction

HCX Disaster Recovery initial Reverse Replication (Failback) completes successfully, however the amount of data transferred is larger than expected.

Environment

VMware HCX

Cause

HCX DR is an asynchronous replication where the changes are:

• written to a primary storage location then copied to a secondary location at a later time. Note there is a delay so that writing to the storage at the primary site does not have to wait for the storage at the secondary site.
• With seeding, the bit-by-bit checksum check that happens can take days to complete on a very large VM.
• The initial full synchronization of the virtual machine disks is a time-consuming process. When vSphere Replication begins to replicate the changed disk blocks (the first incremental sync), the transfer time might require longer than the set RPO time.

 

Example:

• RPO set to 15 minutes
• synchronization starts at 00:00
• it takes five minutes to transfer to the target site
  
    The DR instance becomes available on the target site at 00:05, but it reflects the state of the virtual machine at 00:00. The next synchronization can start no later than 00:10. This replication instance is then available at 00:15 when the first replication instance that started at 00:00 expires.

vSphere Replication

Resolution

This is expected behavior.