Big data and the Cloud that loves it part 2

Part two of of Computerworld Australia's look at big data and more and organisations are viewing the Cloud as their get out of storage hell card.

See part one: Big data and the Cloud that loves it part

Operational data, operational risk

Data stewardship and liability issues are only some of the risks that must be managed when large data sets are moved to the Cloud: as Cloud data becomes more critical to the business, so too does the implicit risk of losing that data to Cloud outages or security breaches that are beyond a company’s control.

Once you put sensitive data in the Cloud, after all, it’s in the Cloud provider’s hands — an issue that hit home recently after a security breach caused hosting provider Distribute.IT to wipe the often-irreplaceable contents of over 4800 websites it was managing.

That incident, and others like it, have kept security front of mind for Saasu, a hosted accounting software-as-a-service provider that stores regular backups of its multi-gigabyte databases — which include customers’ private details and financial information — into the storage Cloud of Macquarie Telecom subsidiary Ninefold.

Incremental backups of Saasu’s 60GB database are replicated to the Cloud every 15 minutes using rSync and Syncified, with a full backup performed every two days and stored in several locations around the world. The total cost? Around $12 per month.

“Keeping data in the Cloud is a no-brainer,” says CTO Paul Glavich. “It’s how our business has been operating for ten years, and we are completely confident in it. Some companies have this touchy-feely thing that if they have the data in front of them, they feel more secure. But it doesn’t make you more secure, and you don’t have the advantage of the large investments these providers have put into it.”

While Saasu is happy with its risk profile, other companies may find their risk profile increasing as they increase their investment in the Cloud.

“The advent of Cloud computing in any enterprise should be used as an opportunity to invigorate enterprise risk management’s links with IT risk management,” says Australian Unity’s Michael. “In addition to technology, Cloud risk should be considered from the perspective of huge and continuing financial investments in IT.”

After weighing these risks, companies will adopt progressively more-complicated models for data storage — then begin feeding Cloud-stored data into Cloud-hosted analytics tools and other applications. This is a significant part of the vision of Cloud platforms like Microsoft’s Azure and Amazon’s combined EC2 application platform and S3 storage. Throw in a commercial-grade repository for value-added data sets in the form of the DataMarket section of Microsoft’s Azure Marketplace, and you’ve got a hybrid Cloud-based model that not only stores your company’s data, but provides easy API-based access to massive databases of useful information at relatively low cost.

DataMarket customers already include geospatial data provider ESRI, credit-scoring data provider Dun & Bradstreet, World Telecommunications/ICT Indicator statistics from the UN’s International Telecommunications Union, and real-estate data clearinghouse Zillo.com.

“Any business running in the Cloud may produce information that’s interesting to track over time,” says Tim Buntel, Azure product manager with Microsoft Australia. “There’s a wealth of information out there but it has been relatively difficult for people to access in the past. Now, companies can put it in the data mart and don’t have to worry about setting up a billing network or distribution network to make sure those large data sets can be easily accessed.”

• Bookmark this page
• Share this article
  •  
  • facebook
  • slashdot
  • digg
  • Reddit
  • stumbleupon
  • linkedin
  • twitter
• Got more on this story? Email Computerworld
• Follow Computerworld on twitter

• Share
  Share this article

  •  google+
  • facebookfacebook
  • slashdotslashdot
  • diggdigg
  • RedditReddit
  • stumbleuponstumbleupon
  • linkedinlinkedin
  • twittertwitter
• print
• email
• Bookmark