I started hearing about Amazon Redshift and I am quite interested in what they have done. As described here they have taken the table structure and flattened all rows to one column. As stated “Using columnar storage, each data block stores values of a single column for multiple rows. As records enter the system, Amazon Redshift transparently converts the data to columnar storage for each of the columns.” . This is a great solution  but what the business needs is a OLTP and OLAP solution where we can replace our OLTP with a hadoop, SAP HANA  or Redshift solution.

The issue being is that all of these solutions give us almost real time but none of them are really a integrated solution where OLTP change = DATAMART change and allow for business driven self service.

Business idea’s have to be tested against current solutions/rules and be promoted to agile teams where the cost of the implementation is estimated. Business may see that the cost will streamline solutions or that current solutions do not warrant a redesign or quick solution. If so all team members have to be brought into a sprint to make the change happen. Automated test must be created if there are more changes to come.

To meet all these needs a system must be, version structured with many versions be it business, developer or production release all in one. Stage able so that any version/part can be rolled back or forward to its parts. and testing is done automatically as defined by the business.

The issue I find with redshift is it seems to be a one way process data into a DataMart that is a OLAP solution but not a total solution.

I have much to learn