The fact that Microsoft has outstripped Oracle in the Data Warehouse and Database Management categories was mentioned several times. Against the background of the BI and analytics market, the SQL Server provides interesting updates with regard to data warehousing.
One of the new features is the Stretch database which enables the users to store a portion of their database in a cloud. This will ensure that older data, known as cold data, can also be accessed easily and quickly. The Stretch database will create a second class database under Azure, and, by making this database stretchable, SQL Server automatically and transparently transfers data into the cloud. This will enable users to submit queries without forcing them to make a distinction between on-premise and in-cloud data sources.
Another feature that aroused interest was the the Column store index. This feature has been developed in order to improve and step up storage, recovery and the management of data. The Column store index enables effective compression of data, enabling it to be transferred faster and facilitating its handling. This will improve cost-effectiveness, opening up new avenues for real-time data warehousing. Using the Column store index, both data processing and data transfer will become faster and more effective. With the aid of the new system, data flows it produces will become easier and more cost-effective to process, and the provision of real-time feedback, intended for the operative systems, will become possible.
A third novelty deserving mentioning is Polybase, a feature which enables the searching and combination by T-SQL of quires from several sources, such as hadoop. This novelty takes advantage of a modern way of storing data when several data sources and technologies are used. Efforts have been made to ensure that the solution will also meet add hoc queries.
When considering the development of data warehousing, its features and the SQL server in a broader context, it can be noted that data warehousing is in the course of transformation and will continue to be so. Today, more is required of data warehousing than the traditional solution DB + ETL + BI can offer. Instead of data warehousing, today’s discussion touches on the ‘data platform’ that comprises several components and serves a number of user roles and needs. The implementation of a modern data platform using an application suite offered by a single supplier is becoming increasingly common, which is reflected in the fact that not only Microsoft but also other suppliers of public cloud services have brought new solutions to the market which enable the acquisition of the entire data platform of a single supplier.