BC/DR in a virtual world

The evolution of business continuity and disaster recovery (BC/DR) has been dramatically changed by the advent of virtualisation. To some extent, the process has been made easier by the independence of Virtual Machines to the underlying hardware; improved mobility allows VMs to easily move between sites and even into clouds. However, BC/DR tools are actually lagging behind the rest of virtualized technology. This issue is highlighted by a recent survey conducted by Zerto that found only 23% of firms are fully confident their DR plan will work with 53% stating that complexity was the biggest challenge they faced in terms of disaster recovery.
As more organisations virtualise servers, storage and networks; BC/DR strategies must be virtualized as well. Today, many business applications, such as SAP or Oracle are managed and provisioned as virtual machines (VMs) and virtual disks (VMDKs) – but current storage-based replication and disaster recovery technologies are still LUN and volume oriented. This adds complexity to BC/DR processes, by requiring IT to replicate entire LUNs rather than specific virtual machines or applications.

Recently, a technology has arisen to solve the problem of disaster recovery in a virtualised environment, namely, hypervisor-based replication. By replicating in the hypervisor or the virtual layer rather than at the storage layer, companies achieve greater efficiency and simplicity, by automating disaster recovery processes.

Simplicity
Hypervisor-based remote replication technology is different from any other DR technology – providing the granularity needed in today’s complex enterprise infrastructures, without compromising on performance, scalability, reporting, mobility or flexibility.

What is a Virtual Protection Group, and how does it help simplify BC/ DR?
Enterprise applications often consist of more than one server. These servers are interdependent so when they are in need of recovery, they must be recovered from a single consistent point-in-time image. This is more complex than it sounds!

Leveraging the concept of storage-based consistency groups that operate strictly at the storage array’s logical unit level, and then interpreting this for the unique capabilities of the virtual platform, has created an important DR innovation: Virtual Protection Groups (VPGs). A VPG is a user-assigned group of virtual machines and their related virtual disks that have dependencies and must be recovered from a consistent image. Today, virtual applications can be recovered in a consistent manner, regardless of where the data resides. This greatly simplifies the job of the IT manager, who previously had to manually build processes to keep applications consistent.

Efficiency
A typical enterprise application includes a web server, application server, database server, etc., and all have their respective disks. Today, administrators tend to put all those disks in a single logical unit in storage so they can replicate the entire application at once without having to search for its individual components. The problem is that this means the entire logical unit must be replicated, even though it may contain other applications that are not in need of replication. That lack of granularity – where administrators cannot identify specific applications and application components to replicate – is inefficient and expensive. Storage-based replication requires a one-to-one ratio of storage, and typically requires matching storage from the same hardware provider on both the production site and the recovery site.

A useful example comes from a credit union in the United States, who began their virtualization initiative in 2008. With virtualization, they were able to quickly provision resources to meet business needs. But they were using a storage-based replication solution that was using more storage than they liked, leading to concerns with the frequency of the replication. By changing to a DR solution replicating in the hypervisor, the company reduced their storage footprint by 43%, and eliminated the need to purchase additional storage for the foreseeable future, which has really helped them reduce costs.

Here in the UK, one of the Europe’s largest retailers with over 988 stores in eight countries, explored existing physical BC/DR tools to protect their data and applications, hoping it would be enough. They evaluated several solutions marketed towards protecting virtualized environments including a popular orchestration tool for virtualized infrastructure, used together with storage-based replication. Unfortunately, the company was still unable to get the agility and flexibility that they needed, namely the aggressive RPOs and RTOs that they required to run a demanding retail business. By choosing hypervisor-based replication the retailer was able to ensure the availability of applications and data and dramatically simplify the migration of environments across Europe.

How does replicating in the hypervisor work?
There are two components used in hypervisor replication:
· An agent-less software module that continuously replicates data from VMs, compressing and sending that data to the remote site over WAN links.
· A management layer which controls replication for the entire virtualized environment and keeps track of applications and information in motion in real time. All management is performed at this level.

Unique to this type of BC/DR is the ability to replicate applications in groups known as Virtual Protection Groups (VPGs). These groups ensure that all the virtual machines (VMs) supporting the mission-critical application are protected consistently. If the DR solution cannot effectively support application groups, ad hoc groupings must be leveraged which can cause errors, especially in high pressure situations.

VPGs ensure that enterprise applications are replicated and recovered with consistency, regardless of the underlying infrastructure. This enables organisations to deploy the application across different physical devices to maximize performance, capacity, and/or to reduce the complexity of the infrastructure.

The benefits of virtualisation are well known and include increased efficiency, flexibility, and savings in space, equipment, and energy costs. However, if you have a virtual or hybrid environment, you cannot realize the full benefits and promise of virtualization unless your replication solution is virtual-aware and virtual-ready. Zerto’s hypervisor-based replication technology is the first and only solution that delivers tier one, enterprise-class, virtual replication and BC/DR capabilities for the data center and the cloud.

Your company’s data will only grow over time; replication demands will also grow, as will the scope of your virtualization initiatives. Purpose-built for the virtual environment, hypervisor-based replication will position you for growth and optimize your business continuity and disaster recovery activities.
www.zerto.com