5 April 2013

Software Application Develoment

With the fat cat “suits” in management always hankering for control, Business Intelligence dashboards may fail to deliver here as very often data is too fragmented to be managed into a small enough number of definable streams. Attempting to combine all data into one ‘logical’ database presents major data formatting, transformation and consistency challenges.

So what to do?

What if you could replicate data from different databases, without custom programming or large-scale integration costs? What if you could share data from one application to another in real time and could be assured that the data would be in the exact format each application and user needed?

OK so data sharing is nothing new and every large vendor worth their salt will have ‘solutions’ in this field. Configuring disparate databases and applications to share their data stores in real-time ranks as one of today’s top IT/business alignment issues.

In an attempt to initiate the hustings on this topic I exchanged words with Alan Arnold who is exec VP and CTO of cluster and data management company Vision Solutions. Arnold told me that traditional methods of data integration rely upon scheduled updates to move data to the centralised data store on a periodic basis.

“This may work for some situations, however these processes can take months to plan and implement. Frequently, portions of the data contained in each of these silos are redundant, but not necessarily in-sync across platforms. This puts the business at risk of making decisions based upon divergent, inaccurate data sources.”

The solution then – sorry, a solution then, or certainly a route to consider in this predicament is a standardised data sharing solution. If you have data integration issues between critical applications that use databases such as Oracle, DB2, Informix, SQL Server, Sybase, or Teradata – then (so says Vision Solutions) a standardised data sharing and replication tool can provide access across the enterprise to accurate, real-time data.

But while a standardised software-based data sharing/replication solution is quick, comparatively inexpensive and (as they say) easy to manage, at its heart it will be database driven rather than programming driven. Because of this the business leaves itself at the mercy of the general direction of open standards. But that’s a good thing right?

Well it is, as long as the business is not so incredibly niche that this just wont work. Say it’s a geothermal analysis company working in offshore liquid natural gas exploration. Click-and-go access to different application silos is not going to sound too strategic at the next IT planning meeting is it?

So I guess it’s each to his own. Standardised data sharing solutions for some and bespoke software application development driven change for others. That’s democratic isn’t it?