By Michele Corvino

In an ideal world, cloud users would never experience downtime or data loss. Providers would keep their infrastructures running flawlessly, 100 percent of the time, with no glitches interrupting business continuity.

But this is the real world. Incidents happen. Whether the cause is a natural disaster, a massive data breach or some malfunction as yet unforeseen, the cloud is not infallible, and sometimes your customers suffer losses.

Unfortunately, infallibility comes at a price. According to IDC, the average total cost of unplanned application downtime per year for the Fortune 1000 is $1.25 billion to $2.5 billion, while the average price tag for infrastructure failure is $100,000 — per hour. Unfortunately, some cloud customers who believe they have comprehensive disaster-recovery (DR) solutions don’t, in fact, have any guarantee. They’ve never tested their plans and have no idea what their recovery times will be. 

One example of this can be found in a study that my organization conducted with ThinkStrategies, in which 10 percent of respondents reported that their DR plans include only a backup. Whether on-site, off-site, tape or disk, a backup plan is not a DR plan. While a backup may be vital to business continuity, it only brings data files back from a single point of time (the last backup) in the event of deleted or corrupt files. In the case of a hardware disruption, organizations need a secondary environment to restore those backups.

An effective disaster-recovery plan needs not only a reliable underlying cloud, but an MSP to help design the customer-specific solution.

In addition, many cloud users fail to test their DR plans, even though it’s the only way to know what will happen in response to a disaster. If the plan doesn’t pan out, it may be too late to save the business from devastating consequences.

This is why it is critical for MSPs to work with their cloud infrastructure partners to encourage customers to have the right DR plans in place and test them regularly before a disaster strikes.

Here are three key areas where MSPs can make a significant impact:

And don’t forget to document. Documenting a DR plan is pivotal to reducing the confusion during a real disaster. A documented practice of a DR plan, known as a runbook, should include the steps, stakeholders, processes and resources needed to recover an environment during testing or in the event of a disaster. It should be updated after each test and after each production change to ensure its effectiveness.

Surprisingly, not all hosting providers help their customers with the documentation process. According to the ThinkStrategies survey, less than 20 percent of respondents stated that they had runbooks provided by their hosting provider — a major hole in the DR safety net.

In conclusion, businesses count on their hosting partners to deliver performance and superior service, and the hosting partner’s success reflects on the MSP. When a customer’s cloud infrastructure goes to hell in a handbasket, they will be looking for help and, just maybe, someone to blame. Will you be the hero or the goat?

Michele Corvino is director of product management at ViaWest.

ViaWest on Twitter: @viawest_inc