The HCX platform uses multiple VMware technologies for virtual machine migration. Migration tech has some overlap with other infrastructure functions, and in some cases may become unavailable for concurrent actions on the same object.

A small example of this is when the Distributed Resource Scheduler kicks in to rebalance a cluster (with intra-cluster vMotion). While a rebalance is happening, HCX cannot initiate a vMotion based site to site relocation operation. a similar concept applies to replications: while a virtual machine is protected using Site Recovery Manager with vSphere Replication, the same virtual machine cannot be also replicated by HCX’s Bulk & RAV migration (which rely on replications) and vice-versa.

We have worked with several customers where migration and business continuity initiatives must co-exist (as much as possible). These are two of the approaches I’ve seen our customers have used to tackle the task while minimizing impact to their continuity plan. Describing these two approaches is the full extent of the post. Something I have had in notes for a bit.

Example Scenario

Virtual Machines in Site A is protected at Site C using Site Recovery Manager with vSphere Replication.

The same virtual machines must be migrated from Site A to Site C.

Method A – By Virtual Machine with HCX vMotion

The concept is that you will use the entire IX appliance and vMotion’s characteristic of saturating the migration resources to trickle in virtual machines. With this approach, you will not remove a vm protection until the virtual machine is fully migrated, and you will not have more than a single workload unprotected.

1. Use HCX vMotion to move a Virtual Machine to Site B.
   
   (Do this only when there’s a completed full initial synchronization to the DR site C in the example scenario)
   
2. After the virtual machine is finished migrating, remove protection. Retain replica disks.
   
   (Stop Replicating Virtual Machine | During step 6 in the linked procedure – select the Retain replica disks check box)
   
3. Re-protect the Virtual Machine from Site B to Site C choosing the retained replica disks as seed.
   

Method B – By Mobility Group with Replication (Bulk/RAV)

The concept with B is to migrate an entire mobility group using any of the available HCX migration types. With this approach, you’re opting to use a maintenance window to remove virtual machine protections for the migration. Temporarily disrupting the desired business continuity state of the virtual machine, to complete the relocation.

1. Slice up the migrations into distinct mobility groups (e.g. 5 – 50 related virtual machines or with similar network patterns). Remove protections for the mobility group opting to retain replica disks.
   
   (Stop Replicating Virtual Machine | During step 6 in the linked procedure – select the Retain replica disks check box)
   
2. Use HCX Bulk or HCX RAV migration to replicate the mobility group.
   
3. Re-protect the mobility group from Site B to Site C choosing the retained replica disks as seed.