Executive overview for your disaster recovery plan:

Fascinating how times change, right? Fifteen years ago, crafting a disaster recovery plan without the cloud meant thinking through every single detail. Now, the cloud takes care of many of those decisions for you. This shift can indeed make a strong business case—showing how traditional plans still hold value but now with the added layer of cloud-based disaster recovery, albeit at an extra cost.

Do you think the added cost is worth the convenience and peace of mind cloud solutions offer today?

Key Requirements for Disaster Recovery

• Datacenter Independent: Ensure the solution is not tied to any single datacenter.

• Fully Redundant Solution: Implement a solution with no single points of failure.

• Fail-over and Fail-back within 48 Hours: Maintain the ability to quickly switch to a backup system and back.

• Portable Solution: Suitable for both hot and cold disaster recovery scenarios.

To optimize replication between the live site and the DR site, compression can serve as a short-term solution but lacks scalability. Block-level replication offers more scalability and the ability to scale up the solution. Using a replication solution specifically designed for DR is crucial, as many products that only offer file synchronization are costlier in DR situations due to higher bandwidth needs.

HP, VMware, Veeam, and DataCore provide robust replication solutions. SAN-level solutions are hardware-dependent and limited to specific vendors. In contrast, a virtual solution is hardware-independent and can operate across diverse hardware, making replication portable and adaptable to various datacenters or service providers offering virtual datacenter services.

The optimal placement for a DR solution is at the hypervisor level, where storage, servers, and networks are all virtualized, creating a virtual datacenter that is hardware-independent. HP, for example, offers a datacenter-in-a-pod solution, which contains all the necessary hardware to support your virtual environment. These virtual datacenters are configured using scripts, policies, and profiles, reducing human error and simplifying the process.

During a DR event, automated scripts and policies can rapidly recreate the datacenter environment in any compatible facility. With two major hypervisors dominating the market, datacenter selection is widely available and can be quickly adapted with minimal configuration. This portability ensures continued operations even if your primary datacenter is unavailable.

To minimize bandwidth usage, implement Exchange DAG with shadow redundancy since snapshot technology is unsupported. Exchange data should be distributed across three offices with a 500ms link, conserving bandwidth for SQL and user data.

This approach is hardware and software vendor-neutral and can be implemented using various providers. For HP-specific solutions, the HP 3PAR offers six-nines reliability, a significant improvement over the standard five nines.

DR Fail-over Plan

• Bonded Warehouse: Automated configuration.

• SLA Performance: Ensure performance standards are met.

• SLA Fail-over and Fail-back: Maintain service level agreements during transitions.

• DR Testing: Regularly test the DR plan.

• Portability: Ensure the DR solution can be moved and implemented easily.

When DR is activated, all users migrate to the DR site. Since the infrastructure is built on policies, scripts, and profiles, setting up a virtual datacenter requires minimal human intervention and expertise, ensuring 24/7 implementation if needed. The solution's agnostic nature means it can operate in nearly any datacenter or cloud provider worldwide, providing flexibility and ensuring that service providers adhere to SLAs, as hosting can be transferred with minimal cost and effort.

Agnostic virtual datacenters represent the future of disaster recovery.