Tuesday, June 21, 2016

The HANA EDW(Enterprise Data warehousing)

Admittingly, the idea of an Enterprise Data Warehouse (EDW) is a mind boggling one. To around, an EDW is basically a DB with loads of information and, nowadays, significantly more because of huge information. Others utilize the terms OLAP and EDW synonymously because of the way that, ordinarily, scientific questions are running on that huge, enormous, over the top measure of information. To a third gathering, an EDW is an idea of architecting, overseeing and administering information from numerous sources, i.e. from numerous, disengaged connections and, along these lines, from various ideas of consistency. In this way, while some think about an EDW as an unadulterated specialized issue (bunches of information, execution, versatility, … ), others take a gander at it as a semantical challenge (single adaptation of reality, consistence, … ). This is the reason some think HANA must be the answer (i.e. to the specialized issue), while others support approaches like BW (Business Warehouse). This online journal will attempt to clarify why both are correct and why SAP is on the way to make what we mark the HANA EDW. 

There is two key test classes to information warehousing: 

One is everything around handling a lot of information, i.e. mass burdens, scientific questioning, table apportioning, versatility, execution and so on. 

The other is around the procedures and the information models inside the information stockroom, i.e. questions around 

Security 

What happens if a section, table or whatever other items is included, changed or expelled? 

What's the effect of those progressions on the (stacking, filing, housekeeping, … ) forms or other, related information models and their hidden inquiries? 

Who has charged something at what minute and why are the outcomes now distinctive? 

What's the relationship between a "client" in table An and the "accomplice" in table B? Is it true that they are the same? On the other hand would they say they are somewhat covering? Are these presumptions ensured by the framework (e.g. through information quality procedures)? 

Has the information been stacked accurately? Have there been any alarms? 

At the point when has the last transfer been performed from which source? 

Are my KPIs (like edge) steady over all my models (and their basic inquiries)? 

The Data Warehousing Quadrant 

A. what's more, B. are fundamentally orthogonal measurements of the information warehousing issue. Truth be told, numerous clients present us their difficulties along those classes. See figure 1 for an illustration. In view of this, figure 2 demonstrates a Data Warehousing Quadrant and sorts frameworks along those two measurements (or "test classifications" as marked previously). The expanding weight to break down the business and its hidden procedures pushes along both measurements – see orange bolts in figure 2: 

More granular information is stacked and broke down. 

More information is accessible through e-trade, sensors and different sources. 

More situations are broke down. 

More situations (information models) are consolidated and incorporated with each other. That requires considerably more exertion declaring uniform expert information, security, consistency, consistence and information quality.