Feeds

back to article Dataupia: a utopian vision for databases?

Over the course of the last few months I have written a couple of times speculating on the development of appliances that might be more generally deployed than the very specific products we have seen to date. In particular, in my last article on this subject I discussed the scope for improving the performance of merchant …

COMMENTS

This topic is closed for new posts.

Threat to Teradata

Dataupia is potentially the biggest threat to Teradata - much more so than the HP NeoVision offering - in that it allows customers to "standardize" on a utility DB like SQL Server and still reap the benefits of an EDW in a heirarchical environment.

The key is the ability to reduce or eliminate porting of existing application - especially Common Off-The-Shelf (COTS) packages and tools that use a standard API to Oracle or SQL Server. Uptake of these tools for Teradata or DB2 has been limited by the number of vendors that have spent the time to add the merchant DB back-ends and custom features to their tools. Dataupia makes this a non-issue.

There are two issues that Dataupia must prove itself in before the threat materializes. First, it must prove that it integrates into a heirarchical environment, where local stores - such as cubes or specialty applications that use a "local" DBMS - can be tied back into the EDW transparently and managed. This is as much an attribute of the vendor database as anything else, but is a key piece to making the use of Dataupia worth while.

Second is the perennial question of scale and complexity. Can Oracle or SQL Server back-ended by Dataupia scale to 100+TB -AND- support the very complex DSS queries and normalized models that merchant DB like Teradata and DB2 support?

If Dataupia simply allows expansion of data by adding more and more denormalized views to the data (and more and more indexes) then it will fail over time simply because it will consume more and more resources to simply stay in one place. If it can allow the existing databases to realize their inherent optimization by better managing the time-consuming aspects of complex joins, then it is a real winner.

Only time will tell - give it two years...

0
0
This topic is closed for new posts.