Building the Data Warehouse

Скачать в pdf «Building the Data Warehouse»


•    performance sensitive


•    accessed a unit at a time


•    transaction driven


•    control of update a major concern in terms of ownership


•    high availability


•    managed in its entirety


•    nonredundancy


•    static structure; variable contents


•    small amount of data used in a process


•    supports day-to-day operations


•    high probability of access



DERIVED DATA/DSS DATA


•    subject oriented


•    summarized, otherwise refined


•    represents values over time, snapshots


•    serves the managerial community


•    is not updated


•    run heuristically


•    requirements for processing not understood a priori


•    completely different life cycle


•    performance relaxed


•    accessed a set at a time


•    analysis driven


•    control of update no issue


•    relaxed availability


•    managed by subsets


•    redundancy is a fact of life


•    flexible structure


•    large amount of data used in a process


•    supports managerial needs


•    low, modest probability of access


Figure 1.9 The whole notion of data is changed in the architected environment.


■■ Primitive data can be updated. Derived data can be recalculated but cannot be directly updated.


■■ Primitive data is primarily current-value data. Derived data is often historical data.


■■ Primitive data is operated on by repetitive procedures. Derived data is operated on by heuristic, nonrepetitive programs and procedures.


■    Operational data is primitive; DSS data is derived.

Скачать в pdf «Building the Data Warehouse»