The most popular and comprehensive Open Source ECM platform
Let us know more
Technology: Bridging the Data Silo
Silos of information have been in the crosshairs of technologists for decades, but it seems like the more people try to eradicate silos, new ones continually pop up. For example, today’s first-generation SaaS applications are providing a fertile ground for creating new data silos, but that is another topic…
The Burton group is out now with a proposal for how to address data silos. It is called the Methodology for Overcoming Data Silos (MODS). It is a process that accepts that data silos aren’t going away anytime soon and that addresses how best to leverage and bridge the data in the silos. And they emphasize the simplicity in their approach.
Lyn Robison, Burton Group VP, said that MODS “does not result in the creation of additional silos. Rather, MODS projects catalog and then clean up the data that resides deep within existing silos so that businesspeople can readily perform cross-silo joins of important enterprise information.”
MODS seeks to form a centralized registry of where data is stored, rather than creating an aggregated new collection of the data itself. Live data is then retrieved, when needed, using Web Services.
The registry provides a way to find where data is housed and to also provide the primary keys for looking up the data in the system where it is managed.
Lyn gives an example in his blog of how UPS is able to unite their many disparate IT systems with the use of the package tracking number. Because all systems understand that same number, data can be aggregated easily real time from very many systems using that single key.
Burton suggests making a list of objects most important to your organization, whether it is a customer, a product, a supplier, a policy, an asset, an employee, a project, a decision, a work activity, whatever. And then make sure that a standard ID is used for that object across all systems. Existing systems may need to be modified to support the idea, but once in place, data sharing becomes a snap.
Web Services here are central to the data retrieval, but the approach is far lighter than standard SOA which bogs down many an organization in implementation.
