Blog Post

Disaster recovery is ripe for cloud disruption

Cloud computing will soon disrupt the market for basic storage and data center backup.

Dropbox, Box and other cloud-based backup tools for desktop and mobile devices have already been wildly successful. Similar tools, such as Riverbed, StorSimple and Ctera, are available for the server market, but they take a hardware-centric approach that has failed to garner a large market.

When Amazon Web Services launched AWS Storage Gateway last month, the move seemed entirely logical, almost expected, given the enormous appetite that desktops have shown for cloud-based data sharing and backup.

But is this enough? Most companies have become pretty good at copying data. The real challenge is not restoring data, but going a level higher and recovering services when disaster strikes. That’s particularly important in the era of IT consumerization, where end users expect to access data through applications of their choice on the device of their choice. In other words, it is useless to protect data and not protect the applications that exploit it.

This is where current backup tools fall short. End users are concerned about how long the outage is going to last, not whether the data is safe (which they assume is a given). Disaster recovery is about minimizing downtime. The cloud has huge potential to make an impact here. By managing and scheduling all of the components involved in service delivery, the cloud could turn recovery time objectives into guarantees.

We are at a fork in the road: backup and disaster recovery are going to be two separate processes.

Backup will be used to quickly restore data — including files, mailboxes, attachments and database tables — and to keep auditors happy about long-term data availability.

Disaster recovery will focus on continuity and service recovery, not data restoration. It will orchestrate all of the components involved in service delivery — from storage to hypervisors, operating systems, databases, middleware and applications — across collaborating clouds. This means that when a cloud disappears, another cloud will be ready to take over at the push of a button. And we’ll be 100 percent certain about the maximum outage time. There will be no need to worry about doing disaster recovery exercises, because clouds will do them on their own, continuously and accurately.

Carlos Escapa is the CEO of VirtualSharp Software. Previously, he was a senior executive at VMware in Europe, where he managed VMware’s field operations in France, Portugal, Spain, Italy and Greece. 

Image courtesy of Flickr user Kevin Burkett.

7 Responses to “Disaster recovery is ripe for cloud disruption”

  1. “backup and disaster recovery are going to be two separate processes: .. umm .. are going to be? They are, and always should be, whether you’re talking onsite or in the Cloud. I do agree that the focus needs to be on recovery time for the restoration of functionality, for that is what enable the business to function.

  2. Sean Hull

    Yep, the potential that cloud computing brings to DR is huge. No need to pay for servers you’re not using until you need them. That said it also introduces new complexity on another level. Push button deploys are nuanced and riddle the ops teams with new challenges.
    As an operational DBA I would say emphatically that it is *NOT* a given that the data is safe.
    I would summarize and say that complex systems fail in complex ways. DR will not be a panacea as this article seems to imply. And I’d be very careful with that word “guarantee”. The devil is very often in the details.

  3. I like this article too. For transparency’s sake, I work for StorSimple, one of the companies mentioned in the article. The terminology of backup, archiving and DR will forever be confused, but the advantages of certain approaches and technologies will not be. Cloud-based DR will revolutionize the DR industry.

    • I’m not sure if my comment was edited after it was first approved or if there was a glitch that removed the 2nd paragraph. Anyway, the part that’s missing now said something like: “software integrated with purpose-built storage has always been an important component of DR and will mostly continue to be.”

  4. Mark Hadfield

    This is an important posting and one that supports our assertions at nScaled. In our opinion, backup has limited value if you cannot recover service to your user community rapidly. Therefore, if you can get rapid service recovery for the same price as backup you should take a hard look at your IT operations. Cloud computing has made cost effective and reliable service restoration a reality for mid sized business with limited budgets. See our expanding client list at to see how the most risk intolerant are able to take advantage of these new realities.