1. Database Structure › 1.1 Design Features

1.1 Design Features

The Database is physically organized using the extensive
SAS database management facilities.  The resultant Database
design consists of a related network of small, segmented SAS
files providing standard physical access to selected portions
of the stored data as well as the ability for relational data
retrieval.  The major features of the Database include:

    o   A single, common method for storing and retrieving
        different data sources (e.g., SMF, CICS, DB2, RMF,
        etc.).

    o   Multi-level data summarization capability to maintain
        data at the detail level as well as summarization to
        the daily, weekly, monthly, and annual levels.

    o   Multi-level database storage providing for online,
        archive (offline), and backup storage of the
        Database.

    o   Efficient processing of large volumes of measurement
        data (e.g., millions of records/day) to allow update
        of the Database from multiple data sources.

    o   Common online inquiry and batch reporting facilities
        to access all data areas of the Database.

    o   Segmentation of data into small, related SAS files
        that minimize the amount of data to be traversed for
        the majority of inquiry/reporting requirements.

    o   Data integrity validation to ensure the accuracy and
        completeness of data maintained.

    o   Backup/recovery facilities to enable restoration of
        part or all of the Database.

    o   Special processing facilities to allow access to
        information not available in the standard online and
        offline Database.

The Database incorporates numerous considerations to
maximize the user's capability to quickly and easily use the
organized information.  These considerations include:

    o  Dictionary defined data elements enabling reference to
       data elements without requiring element format
       definition.  The user need not be concerned about data
       element attributes, thereby saving valuable time in
       inquiry and program definition.

    o  Standard file definitions that provide a common
       definition for a file in terms of data element labels,
       content, and format, regardless of where the file
       resides in the Database.  In other words, the file
       definition for TSO System Activity is identical,
       regardless of whether the request is for a detail,
       daily, monthly, etc., level of the file.

    o  Common retrieval keys are defined in all files to
       facilitate the user's selection, merging, and data
       manipulation.  These common keys include such standard
       elements as system identification, date/time stamps,
       date, time, hour, day, zone, userids, etc.  This
       common key provision enables the integration of unlike
       data sources (e.g., DB2 and RMF) to be easily
       accomplished.

    o  Derived data elements (e.g., average response time)
       are pre-calculated and maintained in the database to
       minimize the user's access time to often-used elements
       of this type.

    o  Data values are carried in decoded form (e.g., device
       type is 3390) and not the typical encoded value
       pattern (e.g., X'80') to improve data usability and
       reduce required analysis time.

This chapter describes how these features have been
incorporated and addressed in the Database.