The 3 Tiers to Building a Disaster Recovery Plan for Your Unique Needs

Gartner estimates the cost of downtime to a business due to an IT disaster to be about $5,600 per minute or $300,000 per hour — and that's just an average — it can be significantly higher1 Direct costs include financial and productivity losses, sales, and operation disruption, legal and investigative costs, credit monitoring and reimbursement costs, extortion payments and settlement costs, and regulatory fines, which doesn't even begin to address the many indirect costs2 Plus, downtime disruptions to your business also puts your brand and reputation at risk, and that cost can be exponential.

Get a Disaster Recovery (DR) plan in place now.

In our highly digital world, workloads are more diverse, mobile, and disparate, across different platforms and geographic regions, whether from the edge, the core, or the cloud. All of it needs to be connected to deliver a consistent digital experience. And, all applications that impact customer and employee experiences must be protected — and available 24/7/365 to meet today's expectations and need for speed. The first step on the journey to IT resilience is to remove systemic risk, which means building a DR plan that ensures your business is protected against any disruption by automating your DR processes.

Think of your enterprise and DR in three tiers.

To build DR resilience, look at your enterprise from a perspective of three different tiers — what's mission-critical and needs to come back up immediately, what's business-critical but can wait a few hours, and what's noncritical and can wait a few days. You can use different methods of protection for different tiers. For instance, multiple terabytes of data from over the past decade may be important to keep for historical or compliance reasons, but it's not going to hurt if it takes several days to recover. Business critical data that is important to your organization's daily functioning may be okay if it takes a few hours to stand back up. But, if it is data that absolutely can't be lost and could be devastating if it's not backed up immediately, it's mission-critical.

1. Noncritical Data

For non-critical data, there are DR solutions that are basically abstraction layers between the on-premise and cloud data centers that allow for migrating VMs from one to the other. One DR use case is the ability to protect workloads at primary or secondary sites. There are lite versions and more robust versions that allow you to live-migrate over distance, with an RPO minimum of five minutes.

2. Business Critical Data

For business-critical data, there are DR solutions that use VM snapshot technology combined with changed block tracking (CBT) to create a replica of a virtual machine. While the snapshot is being taken, a redo file is created, and all changes made during the replication process are written into this file. Once the replication job is finished, the redo file is merged with the snapshot into a live disk file. The next time the replication job is started, only changes made since the last job will be copied to the replica virtual machine. The big advantage is being able to efficiently capture and store data for long-term retention needs. The RPO is also a five-minute minimum.

3. Mission Critical Data

For mission-critical data, the most robust solution is replication technology that deploys small virtual machines on physical hosts. These VMs capture the data as it's written to the host and then send a copy of that data to a replication site. This process results in near-synchronous replication since the data is sent to the DR site at the same time it is written to the production disk array. This replication process is continuous, so the delay between writing data to the host machine and sending it offsite is minimal. The RPO can be measured literally in seconds, so you're not even going to lose five minutes worth of data.

This almost instantaneous replication method is a 100% dedicated disaster recovery solution for virtual architectures with real-time replication orchestration allowing users to do granular failover and DR testing. It has the capability to replicate down to a single VM or at the application level. So, if you're having trouble with a specific database, application, or website, you can orchestrate the failover just for that particular issue, giving you the flexibility to maintain availability and take corrective action with little downtime.

The DR solution that's as unique as you.

OVHcloud® Hosted Private Cloud provides a powerful, private environment dedicated to you — and offers reversibility. You can scale up and down on-demand, with no commitments or hidden charges for network usage and no ingress, egress, or API call fees.

The experts at OVHcloud can help you determine the best DR solution based on your unique IT enterprise and business operations. Learn more at ovhcloud.com.

1 - Gartner Blog - "The Cost of Downtime"
2 - Study – "Issues in Information Systems"