How Zerto Recovery Works

Installing Zerto Virtual Replication installs the Zerto Virtual Manager, which sits in the hypervisor layer on the enterprise site. You manage the Zerto Virtual Manager, one on each of the protected and recovery sites, using the Zerto User Interface.

Zerto also provides a set of RESTful APIs and PowerShell cmdlets to enable incorporating some of the disaster recovery functionality within scripts or programs.

In the protected site you define the virtual machines that you want to replicate, either individually or together, as a virtual protection group (VPG). The virtual machines that you include in the VPG can come from one or more hypervisor hosts. In this way, you can protect applications that run on multiple virtual machines and disks as a single unit – a VPG. An example of an application that runs on multiple virtual machines includes software that requires a web server and database, both of which run on virtual machines different than the virtual machine where the application software runs.

A virtual machine can be included in several VPGs so that you can recover it to several sites, depending on the needs of the organization. For example the same workload can be protected to a local or a remote location as well as to the cloud. Using several recovery sites also enables migrating disaster recovery datacenters from one location to another.

Note:

CD and DVD drives cannot be protected.

Every write is copied by Zerto and sent, asynchronously, to the recovery site, while the write continues to be processed on the protected site. For greater efficiency and performance, the write can be compressed before being sent to the recovery site with throttling techniques being used to prioritize network traffic.

On the recovery site the write is written to a journal managed by a Virtual Replication Appliance (VRA). Each protected virtual machine has its own journal. Every few seconds, a checkpoint is also written to each journal. These checkpoints ensure write order fidelity and crash-consistency to each checkpoint. During recovery you pick one of these crash-consistent checkpoints and recover to this point. Additionally, checkpoints can be manually added by the administrator, with a description of the checkpoint. For example, when an event is going to take place that might result in the need to perform a recovery, you can pinpoint when this event occurs as a checkpoint written to each journal.

The VRA manages the journals for every virtual machine that will be recovered to the hypervisor hosting that VRA. It also manages images of the protected volumes for these virtual machines. During a failover, you can specify that you want to recover the virtual machines in the VPG using the last checkpoint or you can specify an earlier checkpoint, in which case the recovery of the mirror images under the VRA are synchronized to this checkpoint. Thus, you can recover the environment to the point before any corruption and ignore later writes in the journal that were corrupted, regardless of the cause of the corruption, such as a crash in the protected site or a virus attack.

To improve the RTO during recovery, the user is able to start working even before the virtual machine volumes on the recovery site have been fully synchronized. Every request is analyzed and the response returned either from the virtual machine directly or from the journal if the information in the journal is more up-to-date. This continues until the recovery site virtual environment is fully synchronized, up until the last checkpoint or an earlier checkpoint, when the integrity of the protected site was assured.

When you need recovery to a point further in the past than the time saved in the journal, you can restore from files created during the Retention process. Retention is an extension of disaster recovery, with the virtual machine files, such as the configuration and virtual disk files, saved to a repository . These files are then used to restore the virtual machines to the point of the stored retention sets at the recovery site.