inicio mail me! sindicaci;ón

Managing Data in the Clouds

by Joe McKendrick

Will integration issues dampen the enthusiasm around cloud computing? What are the role of data environments in these new scenarios?

The folks that manage data integration have some interesting observations to make on this topic. I recently had the opportunity to speak with Chris Boorman, chief marketing officer with Informatica, and Ron Papas, senior vice president and general manager for Informatica On Demand, about the enterprise data management implications of this growing trend.  (The interviews are posted here and here in this two-part series.)

As cloud computing engagements increase in sophistication and edge ever closer to the mission-critical core of the enterprise, recognition is growing that there are enterprise data management issues that still need to be worked out. “Our belief is that cloud computing or on-demand computing is simply a way of further fragmenting data, because customers are absolving themselves from responsibility for the management, storage, security, and backup and recovery of the availability of that data,” Chris pointed out. However, he emphasized, “you must never, ever, absolve responsibility for the quality and the ownership of the data, and having such quality and ownership as part of your core business processes. And that requires integration.”

Cloud computing hands off many of the aches and pains associated with systems and application development and management to someone else. But this does not relieve enterprises of the requirements and responsibilities around effectively managing enterprise data. Many observers, in fact, are concerned about the implications of cloud computing on enterprise data management and integration, since much of the processing and storage of information shifts to outside providers.

As Informatica’s Ron Papas put it, technically, there isn’t a lot of difference between on-site systems and data stores and cloud-managed systems and data stores.  However, there’s a big difference in the ownership of these applications:

“What’s that’s doing is it’s bypassing the traditional process of having IT design the whole integration processes into the solution. So, before you know it, you could be up and running with Salesforce.com without having put much thought into integration, because it’s really being led by the line of business side. You could have someone in the sales and marketing unit that somehow bypassed IT and went up and implemented Salesforce. All of a sudden, they realize they need access to that data. they need it synchronized.”

More companies are emphasizing their ability to compete on analytics, and the ability to integrate and leverage enterprise data is key. Whether on-site or in the cloud, effective data integration is a must.

Share and Enjoy: These icons link to social bookmarking sites where readers can share and discover new web pages.
  • E-mail this story to a friend!
  • Print this article!
  • TwitThis
  • del.icio.us
  • Facebook
  • Reddit
  • bodytext
  • Google
  • StumbleUpon
  • SphereIt


2 Comments »

ToddAugust 22nd, 2008 at 1:16 pm

Not having reviewed the interviews themselves –

Ownership and accountability — “that requires integration” — traditional integration? Or can the combination of syndication and semantics bridge the cloud?

The inherent issue with traditional lineage and control is that ETL and warehouse vendors still think 1 model, 1 repository is required to achieve organizational data visibility. In some cases, the marketing language even centers around how many connectors are available. Connectors that simply reposition data (and yes shuttling data around the organization and across cloud boundaries is painful - especially when you don’t control the services).

What I am wondering is who is thinking about actively infusing integration technologies with new constructs (such as semantic models) that relish the thought of silos and cloud-based services across the enterprise while still maintaing analytic capabilities?

Ike JohnsonAugust 22nd, 2008 at 3:22 pm

I wish reporters would stop publishing marketing non-sense from marketing people who are promoting expensive proprietary software; these folks are clearly not practitioners.

Cloud computing and SaaS platforms are a reality today and they are simple to integrate with open source platforms like Jitterbit and others.

Data Ownership is not an issue with cloud computing; your organization owns its data and any SaaS offering that’s worth using (like Salesforce.com) ensures that access can be strictly controlled. The organization that pays the SaaS vendor will be the owner of that data; he who pays owns - there’s no other way.

Organizations simply need to determine how best to run their business operations and if they need their internal systems to talk to their applications in the cloud then they should be using open source technologies, that are readily available, to design and integrate their business processes.

The vision where there is some central control over an entire enterprise’s data access makes for great white papers, but is simply not implemented across an entire organization; it’s been touted for years and is never a reality.

Different groups within an organization own their applications and when they need to interact with other groups to improve their processes they have tons of options for how to get the work done.

Your comment

Want an image to appear near your comment? Go to gravatar.com

HTML-Tags:
<a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>