Zerto: replication and recovery in DRC

Zerto is an enterprise continuous data protection solution capable of restoring virtualized applications in public, hybrid and private cloud environments. It is an HPE enterprise-class solution that enables:

• Replicate virtual machines (VMs) in real time with minimal performance impact. Replicate between sites and within a single site;
• Automatic failover when a failure is detected;
• Testing Disaster Recovery (DR) scenarios without affecting the production environment. As part of the scenarios, we determine the order in which the systems will start up, and indicate the addresses at which the systems in question should automatically start up;
• Restore data from any point in time thanks to a journal-based recovery mechanism that continuously executes replication points (checkpoints) to which we can perform restoration;
• Support for multiple virtualization and cloud platforms, making it a flexible solution for different IT environments.

Zerto is distinguished by the fact that it does not require agents installed at the operating system (OS) level of individual VMs, which simplifies deployment and management.

The standard Disaster Recovery Center architecture from Zerto includes several basic elements.

Production site – the main location where production systems run and from which data is replicated.

Backup site (DRC) – the location to which data is replicated and where services can be launched in case of failure.

Zerto Virtual Manager (ZVM) – a replication management component that integrates with vCenter or SCVMM. Each site has its own ZVM. To ensure data replication, the ZVM data must see each other. From one ZVM, the other connected ZVM can be managed (more management convenience).

Virtual Replication Appliance (VRA) – a vitrual machine invoked from the ZVM level on each selected hypervisor host, responsible for transferring replicated data. Along with the VRA, a VRA-H VM is created automatically, which supports the VRA VM in operation (both must always work together and cannot be moved). The architecture assumes correct operation of replication only if the hypervisor of the replicated source VM and the hypervisor designated as the target for the replicated VM have dedicated VM VRAs running and operational. If you migrate the VRA to another hypervisor or move the replicated VM to a hypervisor without a running dedicated VRA, replication will stop.

Zerto Journal – a mechanism that allows us to recover data from a selected point in time, even from a few days ago, which allows us to respond to ransomware attacks. The size of Journal affects how far back to a given point we can go. Journal is located by default in the directory of the replicated VM. A startup allocates 16 GB and its size is dynamically changed within the journal history parameter set on a time scale. The limit at which replication is stopped is less than 30 GB of free Datastore space or less than 15% of free Datastore space. When stopped, the historical data (recent synchronizations) is not deleted – the RPO increases. When the data space is increased, replication resumes. Journal sizing is recommended to be done based on the variability of the replicated VM data.

This architecture ensures full redundancy and the ability to quickly switch services in case of failure.

The process of implementing a Disaster Recovery Center using the Zerto solution is divided into several key stages:

• Analysis of the IT environment – identifying critical systems, applications and data that must be protected;
• DR architecture design – selection of DRC locations, determination of RPO/RTO requirements, selection of hardware and network resources;
• Installation of Zerto – implementation of ZVM and VRA on both sites, configuration of network connections and accesses;
• Replication Configuration – Create Virtual Protection Groups (VPGs), which are logical groups of VMs that are replicated together. VPGs are used in DRC scenario planning;
• Testing DR scenarios – conducting failover and failback tests, verifying compliance with security policies;
• Monitoring and optimization – ongoing tracking of performance, alerts, SLA compliance, and configuration updates as the environment changes.

The use of Zerto in the construction of the DRC succeeds in minimizing the RPO (Recovery Point Objective) and RTO (Recovery Time Objective). Data is recovered within minutes, significantly reducing system unavailability time.

VM replication executes regardless of the state of the VM (running/off), plus there is integration with orchestration systems. Such automation greatly simplifies management.

The scalability of the solution makes it easy to expand the environment as the organization’s needs grow.

The solution is compliant with generally recognized regulations and standards, including industry standards, such as ISO/IEC 27001, ISO/IEC 27017, NIS2. On-the-fly and at-rest data encryption is provided, supporting the cyber security of data and systems.

Cloud integration, including the ability to replicate to Azure, AWS, Google Cloud increases the flexibility and availability of the backup infrastructure.

It is worth noting the ease of recovering individual data – it is possible to restore individual files by File-Level Recovery (FLR) from individual protected checkpoints. Replication management is flexible – you can reverse the direction of running replications.

Building a DRC using Zerto is an investment in an organization’s resilience to failures and incidents. With its advanced replication, flexibility and management simplicity, Zerto is the foundation of a modern Disaster Recovery strategy. Organizations that implement such a solution gain not only data security, but also a competitive advantage through business continuity.