StarWind is a hyperconverged (HCI) vendor with focus on Enterprise ROBO, SMB & Edge

VMware vSphere Replication 8.4 details and enhancements

  • March 18, 2021
  • 10 min read
IT and Virtualization Consultant. Vladan is the founder, and executive editor of the ESX Virtualization Blog at He is a VMware VCAP-DCA and VCAP-DCD, and has been a vExpert from 2009 to 2023.
IT and Virtualization Consultant. Vladan is the founder, and executive editor of the ESX Virtualization Blog at He is a VMware VCAP-DCA and VCAP-DCD, and has been a vExpert from 2009 to 2023.

VMware has recently released new version of their popular replication software vSphere Replication 8.4. This software is bundled with vSphere for free, however if you want to do some orchestration and automation, recovery plans with automated tests, you’ll need to buy VMware Site Recovery Manager (SRM).

The latest release of vSphere Replication (VR) 8.4 is finally available with international languages. Yes, VMware has added localization, finally.

The software is available in English, French, German, Japanese, Korean, Spanish, Simplified Chinese and traditional Chinese languages.

If you don’t know vSphere Replication, here is a small schema. The overall architecture of vSphere replication is pretty simple as you can see. We can replicate our VMs from one site to another with as low as 15 min interval.

Note: While there is a possibility to get even lower RTO (5min) this adds some restrictions. You can use the 5 minute Recovery Point Objective (RPO) if the target and the source sites use VMFS 6.0, VMFS 5.x, NFS 4.1, NFS 3, VVOL and Virtual SAN 6.0 storage and later. The 5 minute RPO can be applied to a maximum of 100 VMs on VMFS 6.0, VMFS 5.x, NFS 4.1, NFS 3 and Virtual SAN 6.0 storage and later. The maximum for VVOL datastore is 50 VMs.

The underlying storage device does not matter. It can be SAN, NAS or VMware vSAN or StarWind VSAN. From any storage to any storage.

Basically, the software works at the VM level, not at the storage level. It replicates VM’s disks to other location where those VMs are in standby.

VMware vSphere Replication

VMware vSphere Replication

What’s New in vSphere Replication 8.4?

  • Reprotect Optimization – VR is meant to be used with SRM, for easier automation. The new Reprotect optimization when VR is used with SRM, there is the Checksum which that is no longer used for reprotect run soon after a planned migration (was lengthy). Instead, changes are tracked at the site of the recovered VM and only those changes are replicated when reprotect is run.
  • New config service/management UI.
  • Support for vSphere Native Key Provider – VRMS Appliance Management Interface.
  • Support for virtual NVMe controller.
  • Ability to select multiple VMs to move between vSphere Replication servers and to reconfigure replications.
  • vSphere Replication 8.4 continues to release accessibility enhancements based on VPAT (Voluntary Product Accessibility Template) tests

Note: There is a few caveats, especially for users which upgraded their vSphere version to vSphere 7 Update 2. It touches encryption and support bundles. Here are some details about that.

  • NEW vSphere Replication 8.4 does not support vSphere 7.0 Update 2 if virtual machine encryption is switched on.
  • NEW vSphere Replication 8.4 does not provide support bundle management in the VRMS Appliance Management Interface. This includes lists with support bundles and deleting support bundles. To manage the support bundles through SSH, establish an SSH connection to the vSphere Replication Appliance.

You can check the full release notes here.

vSphere replication architecture and functions

Basic functionality of VR is this. vSphere Replication agent sends changed blocks in the virtual machine disks from the source site to the target site. The changed blocks are applied to the copy of the virtual machine. This process Is completely independent from the storage layer.

During replication configuration, you can set a recovery point objective (RPO) and enable retention of instances from multiple points in time (MPIT).

While replication from one site to another is the simplest and most common scenario, other scenarios and architectures are very interesting too. For example, on the image below you can see that we can perfectly replicate from two different main sites to third, independent site where our “sleeping” VMs reside.

In case we have a problem on one of our main sites, we can initiate a failover to the third site and production can continue. We only lose time that we have set on our RTO.

Replication from two sites to single target site

Replication from two sites to single target site

The network replication traffic can compress and encrypted. But if you enable compression, this will need more CPU resources on both source and destination datacenter.

If you have a lot of VMs to replicate, VMware VR can scale out by adding additional VR servers to the overall architecture. It is at no cost.

vSphere Replication 8.4 Limitations

We have talked about the 5 min RTO and some of the limitations on the datastore. Now let’s talk about the limitation to each VR server. (yes, there are some).

The maximum number of virtual machines that you can protect by using vSphere Replication lower than version 8.4 is 2000 per vSphere Replication appliance.

vSphere Replication 8.4 and newer can protect up to 3000 virtual machines alltogether. The maximum number of virtual machines that you can assign to each vSphere Replication server is 200. So, if you deploy the maximum of 9 additional vSphere Replication servers, the total number of virtual machines that you can protect is 2000. This is pretty good, even for larger infrastructures.

Final Words

VMware VR evolves and stays integral part of vSphere. It’s kind of a product that is available for free from VMware as it has the basic function to enable and operate VM replication to remote locations. If you’re using backup software for your VMs, most of the vendors allows you to setup replication of VMs from within the backup software itself without the need of deploying additional VMware product to your installation.

However, if you want to leverage automation and SRM, then VR might be the way to go. VMware does a good job by improving and enhancing the solution every year.

Hey! Found Vladan’s article helpful? Looking to deploy a new, easy-to-manage, and cost-effective hyperconverged infrastructure?
Alex Bykovskyi
Alex Bykovskyi StarWind Virtual HCI Appliance Product Manager
Well, we can help you with this one! Building a new hyperconverged environment is a breeze with StarWind Virtual HCI Appliance (VHCA). It’s a complete hyperconverged infrastructure solution that combines hypervisor (vSphere, Hyper-V, Proxmox, or our custom version of KVM), software-defined storage (StarWind VSAN), and streamlined management tools. Interested in diving deeper into VHCA’s capabilities and features? Book your StarWind Virtual HCI Appliance demo today!