In a Zerto environment, a Virtual Protection Group (VPG) is the fundamental unit of management used to group virtual machines that must be replicated together to maintain write-order fidelity and application consistency. This is particularly vital for multi-tier applications, such as a database server and a web server, that need to be recovered to the exact same point in time.
According to the HPE Advanced Storage Solutions technical guides and Zerto's architectural specifications, a single Virtual Machine (VM) can be associated with a maximum of three VPGs simultaneously. This capability is often referred to as "one-to-many" replication. This architectural flexibility allows a storage administrator to design complex data protection strategies that go beyond simple site-to-site disaster recovery. For example, a VM could be part of:
A Local VPG for high-speed recovery from the local journal (Short-term retention).
A Remote VPG for disaster recovery to a secondary data center or public cloud.
A Tertiary VPG for long-term retention or to a third site for regional disaster protection.
When a VM is protected in multiple VPGs, each VPG maintains its own independent journal, settings, and Recovery Point Objective (RPO) targets. However, the Virtual Replication Appliance (VRA) on the host only needs to read the data changes (IOs) from the hypervisor once; it then distributes those changes to all the target VRAs associated with the various VPGs. This ensures that while the VM is highly protected across multiple locations, the overhead on the production host and the hypervisor remains minimal. It is important to note that while three is the maximum, the storage architect must ensure that the available network bandwidth and the IOPS of the target storage systems can handle the aggregate replication load of all associated VPGs.
