Defining an Analytic Workspace

How to define an analytic workspace

Analytic workspaces are defined using commands in the OLAP DML. There are two methods by which this can be accomplished:

• Use the OLAP API's SPLExecutor method to issue OLAP DML commands. This allows applications using the OLAP API to create new analytic workspaces and alter existing workspaces. When workspaces are defined through the SPLExecutor method, they can be temporary (that is, for the life of the session) or they may be persisted.

• Use OLAP Worksheet to issue OLAP DML commands. OLAP Worksheet connects to an OLAP Services instance as a new session and allows you to work interactively with the OLAP DML; this is similar to issuing SQL commands from within SQL Worksheet. When an analytic workspace is created using OLAP Worksheet, it must be persisted so that OLAP Services applications using the OLAP API will be able access it.

This guide discusses how to use OLAP DML commands to define an analytic workspace.

Examples: Defining an analytic workspace

The following example creates a new analytic workspace named shoes; the full name of the new analytic workspace is shoes.db.

database create shoes

The following example creates the shoes.db analytic workspace in a directory named apps on the i drive of an NT system.

database create 'i:/apps/shoes'

For the complete syntax for the DATABASE command, see the OLAP DML Reference.

About the term "database"

Throughout this guide, you will notice that the OLAP DML command `database' is used to create and manage analytic workspaces. When referring to the OLAP DML, you can think of the terms `database' and `analytic workspace' as being equivalent. The `database' command is used in the OLAP DML to allow for compatibility with the Express Server stored procedure language. (Express Server was the predecessor to OLAP Services.)

Do not confuse analytic workspaces with the Oracle relational database. Analytic workspaces are stored in files that are separate from Oracle relational database files.

Managing analytic workspace structure and size

An analytic workspace can be made up of many files. There is always a main analytic workspace file. There can also be one or more extension analytic workspace files. You can use extension files to divide a single analytic workspace among several files, so the analytic workspace can be larger than the space that is available on any single disk. Typically, you need extension files only when the analytic workspace is located on a disk with limited available space or when the analytic workspace will grow to a very large size. An analytic workspace that is stored in more than one file is called a multifile analytic workspace.

When you use the DATABASE command with the CREATE keyword, a new analytic workspace file is created. As the analytic workspace is populated, data is added to that file and, optionally, additional analytic workspace extension files are created, if needed. Depending on the options that you specify when you create an analytic workspace, you can change the default characteristics of these files:

• The maximum size of analytic workspace files

• The increment size of analytic workspace files

• The location of the main analytic workspace file

• The location of analytic workspace extension files

  Note: If you want to specify location of analytic workspace extension files only for a given session, then use the DBEXTENDPATH option. If you want to specify the location of analytic workspace extension files only for an instance of OLAP Services, then use the ExtensionFilePath setting of the OLAP Services Instance Manager.