This content is part of the Essential Guide: VMware virtual recovery and backup best practices and tools

How do you use VMware replication for disaster recovery?

Admins can protect their data centers from disaster by replicating them to another location with vSphere Replication, which supports replication across sites and clusters.

VMware uses hypervisor snapshot technology to enable the replication of VMs from one site to another. VMware replication...

is a simple disaster recovery strategy that can prevent data loss, and with tools like vSphere Replication, it isn't difficult to do.

VSphere Replication enables rapid VM restoration

VMware replicates VMs with a built-in product called vSphere Replication that can replicate a running VM to another location. VMware replication creates a copy of a VM that can be stored within a cluster locally or at another site; this provides a data source for the rapid restoration of a VM.

VSphere Replication has a management UI and a virtual replication appliance. The appliance can be installed at the local site and the remote site.

The virtual replication appliance pushes the data from one site and receives it at the remote site. VSphere Replication copies virtual machine disk files -- which track VM changes -- and ensures that changed blocks are replicated to the remote site.

VSphere Replication establishes a recovery point objective (RPO), which determines the age of files that must be recovered from backup in order for the system to return to normal operations. The RPO can be as low as 15 minutes, such that if a disaster were to occur at the main site, you'd only lose a maximum of 15 minutes of your work.

VMware assumes that you have a second vCenter server installed at the remote site. If you don't, vSphere Replication also supports the replication of VMs from source to target site, from source cluster to destination cluster and from multiple source sites to a shared remote target site.

Replication can save you from disaster

You can set up VMware replication with the destination as your remote office -- a remote ESXi or vSphere cluster. After the first full copy, the original VM is registered on the destination ESXi host in a powered-off state.

If, for instance, you set up an RPO of four hours, VMware replication will automatically replicate changes every four hours. The network traffic isn't affected as only incremental changes to the source VM are replicated.

If you have a disaster in your main data center, you can connect to your remote site and power on the replicated VMs.

If you have a disaster in your main data center, you can connect to your remote site and power on the replicated VMs.

To perform a failback operation, log in to the target site and manually configure a new replication in reverse, from the target site to the source site. VSphere Replication uses the disks on the source site as replication seeds so that it synchronizes only the changes made to the disk files on the target site.

VMware vSphere Replication does not have orchestration for automatic failover or failback. VMware has another product for this purpose called Site Recovery Manager.

Many backup software vendors incorporate backup features with VM replication. These products use VM replication to protect VMs that run in the main data center and replicate them to a remote data center by using the same underlying hypervisor snapshot technology.

Some backup products also integrate orchestration for fast recovery and can plan the correct order for tiered application workloads with the database VM first and the web front end second.

Next Steps

Learn how to copy VMs with vSphere Replication

Compare vSphere Replication to the competition

Use logs to troubleshoot vSphere Replication issues

Dig Deeper on VMware backup software