ETL Rule based engine

karthi_gana · Post by **karthi_gana** » Thu Sep 13, 2012 1:37 am

All,

I have searched about this topic and found the below one which is saying datastage is very strong in 'T' and also datastage is not that much strong in Rule based. This discussion happened 7 or 8 years back. But the things are changed now. we are in Datastage 8.7 version. How about the latest version i.e 8.7 or 8.5?

viewtopic.php?t=87803&postdays=0&postorder=asc&start=0

ray.wurlod · Post by **ray.wurlod** » Thu Sep 13, 2012 5:34 am

DataStage now (8.7) has a Data Rules stage, so I guess you can say that it's now strong in Rules.

Whatever that may mean.

Mike · Post by **Mike** » Thu Sep 13, 2012 6:03 am

Certainly a bit more functionality in 8.7, but still far short of a comprehensive Business Rules Management System (BRMS).

Check out the IBM ILOG product at http://www-01.ibm.com/software/websphere/ilog/.

DataStage interfaces nicely with ILOG via java or web services.

Mike

vmcburney · Post by **vmcburney** » Thu Sep 13, 2012 6:50 pm

There is now a good interface for data quality rules, these are mostly rules that require a boolean result. The Data Quality Rules stage makes it easier to re-use rules across DataStage jobs and govern changes to those rules over time.

For more wide ranging collections of rules it is hard to beat iLog from the WebSphere brand - though it requires data in an XML or java format and this could have implications on performance in batch environments.

You also have custom parallel routines written in C++ for high performance rules or whatever you can write in Transformer functions.