| A data warehouse configuration is commonly known | | | | typically stored in the data warehouse. In other words, |
| as the logical architecture. It is the foundation on which | | | | an ODS stores only very recent information, rather |
| a data warehouse is built or you can say the logical | | | | than storing permanent information like the data |
| architecture is a configuration map of the data to be | | | | warehouse. |
| stored in the warehouse. A data warehouse | | | | - Individual Data Mart: It is the summarized subset of |
| configuration includes a central Enterprise Data Store; | | | | the enterprise's data specific to a functional area or |
| an optional Operational Data Store; one or more | | | | department, geographical region, or time period. |
| optional individual business area Data Marts; and one or | | | | Access to the EDS can often be difficult or slow, |
| more Metadata Store or Repositories. When | | | | thanks to the volume of data it contains. This is when |
| discussing options with your data warehousing | | | | the role of Data Marts comes into play. Data Marts |
| consultant, these are some of the things that you must | | | | filter, condense and summarize information for specific |
| consider. | | | | business areas. However in the absence of the Data |
| - Enterprise Data Store (EDS): This is the central | | | | Mart layer, users can access the EDS directly. |
| repository that supplies atomic integrated information | | | | - Metadata Store or Repository: This is a catalog of |
| to the whole organization. The EDS can be defined as | | | | reference information about the primary data. It |
| the cornerstone of a data warehouse and this can be | | | | provides users and developers with a road map to the |
| accessed for immediate informational needs as well | | | | information in the data warehouse. Metadata is further |
| as for analytical processing to assist strategic | | | | divided into two categories- information for technical |
| decision-making in an organization. The EDS may | | | | use or transformational and information for business |
| contain data from the existing subject area operational | | | | end-users. Transformational metadata serves a |
| systems as well as from external sources. This data | | | | technical purpose for development and maintenance |
| further feeds individual Data Marts that are accessed | | | | of the warehouse, while end-user metadata serves a |
| by end-user query tools at the user's desktop. The | | | | business purpose. Basically, this type of metadata |
| EDS can also be called as the collection of daily | | | | translates a cryptic name code that represents a data |
| "snapshots" of enterprise-wide data taken over an | | | | element into a meaningful description so that end-users |
| extended time period. Thus it creates an optimum | | | | can recognize and use the data. On the other hand, |
| environment for strategic analysis. | | | | transformational metadata maps the data element |
| - Operational Data Store: It is a "snapshot" of a | | | | from its source system to the data warehouse, |
| moment in time's enterprise-wide data. An ODS is a | | | | identifying it by source field name, destination field code, |
| set of relational databases designed to perform simple | | | | transformation routine, business rules for usage and |
| queries on small amounts of data as opposed to the | | | | derivation, format, key, size, index and other relevant |
| complex queries on much greater volumes of data | | | | transformational and structural information. |