To prepare the environment for the migration (including on-premises vSphere, target SDDC on VMware Cloud on AWS, SQL Server Instance configured as FCI), use the following steps. Some of them might require a separate action plan, which is outside of the topic of this publication.

Step 1. Identify VMs

Step 2. Record the shared disk configuration

Step 3 Record the network configuration

Step 4  Prepare SQL Server for maintenance

Step 5 Take a backup

Step 6 Record cluster configuration

Step 7 Configure Hybrid Linked Mode

Step 8 Remove DRS and HA overrides

Step 9 Check virtual Compatibility

Step 10 Prepare the SDDC

Step 11 Execute a test vMotion

 

Step 1. Identify VMs to be migrated and place them into a separate vSphere Folder:

Step 2. Take notes of the VM shared disk configuration:

 

Step 3. Take notes of the VMs network configuration. Ensure that all routed cluster networks used by cluster nodes are stretched (L2 VPN or HCX) and available in the target SDDC. Non-routed networks (if present) should not be stretched and should be pre-created in your SDDC.

Step 4. Prepare all applications using databases on the cluster and SQL Server instance(s) for the maintenance. Ensure that the cluster has no errors or warnings in the cluster.log or Windows Event Log.

Step 5. Take a full backup of the SQL Server data located on disk drives hosted on shared disks (pRDMs). Store the backup outside of virtual machine(s) to be migrated.

Step 6. Take notes of the cluster configuration using the Failover Cluster Manager mmc, including the quorum configuration

And network configuration

Step 7. Ensure that Hybrid Linked Mode is configured between your on-premises and VMware Cloud on AWS environments and you are able to migrate VMs using vMotion from on-premises to the cloud.

NOTE: HCX could not be used to move a VM with pRDMs attached to VMware Cloud on AWS at this time. The first node of a cluster should be migrated manually using vSphere Client. All other nodes can be migrated with HCX (in this case, for all VMs to be migrated with HCX, pRDMs pointers should be detached and vSCSI controller bus sharing should be changed to none before the migration). If the migration of the first node without HCX is not an option, engage VMware PSO for a possible workaround.

Step 8. Remove all VMs to be migrated from all DRS Rules and Groups (when all VMs, members of a  DRS Group should be removed, the DRS Group must be removed as well). Take notes of the DRS configurations

Attempting to migrate a VM being part of DRS Rules or Groups would fail with the error: “permission to perform this operation was denied”:

Step 9. The virtual Compatibility  vHardware version ) for VMs should be at least v11. Upgrade if required

Step 10. Prepare the target SDDC

  1. Create a separate Virtual Machine Folder to place migrated VMs

    2. Precreate a vSAN Storage policy for pRDMs to be migrated. The following settings are recommended for performance:

  • RAID1 for performance.
  • FTT=1 or higher to ensure data redundancy

  • Use space reservation – thick provisioning to avoid out of disk space situations

3. Check that all required cluster networks are stretched to the cloud and available. Strech network segments via L2VPN or HCX if required.

Step 11. Execute a test vMotion of a test VM  to check the relocation time and the network connectivity. Take notes of the time used to relocate data and use the estimate for the migration of the production workload.

check-circle-line exclamation-circle-line close-line
Scroll to top icon