Planning Your Data Stores and Processes

Jobs

After using the Explorer to define the properties of a given data store, you will use the Process Editor to define its Job. A Job is a metadata record that specifies the processes that create one or more data stores. In Job and Process Flow in the Process Editor, a Job that creates the Customer Detail Table is represented by the icon with the rectangle around it in the left panel of the Process Editor.

Job and Process Flow in the Process Editor

[Job and Process Flow in the Process Editor]

To create the Customer table, you would click the Customer Job with the left mouse button and select Run from the pop-up menu. The Load Generation/Execution Properties window would display. From the Load Generation/Execution Properties window, you could execute the code associated with the Customer Job.

If you want SAS/Warehouse Administrator to generate the code for a Job,

the Job's SAS/Warehouse Administrator Generated attribute must be set (it is set by default)
you must create a Process Flow for the Job in the right panel of the Process Editor.

If the Job's SAS/Warehouse Administrator Generated attribute is set, SAS/Warehouse Administrator generates code for each process and data store in the Process Flow, from the bottom to the top.

If you want to specify your own code for a Job,

the Job's User Written attribute must be set
the Job must specify the location of the user-written source code to be executed for this Job.

If the Job's User Written attribute is set, SAS/Warehouse Administrator retrieves code that creates the data store(s) for that Job. If you will supply the source code for a Job, no Process Flow is required, although you might want to create one for documentation purposes.

A Job can include scheduling metadata that enables the Process Flow or user-supplied program to be executed in batch mode at a specified date and time. The steps for creating Jobs are described in Maintaining Jobs.

Job Inputs and Outputs

Keep the following in mind as you plan Jobs for which SAS/Warehouse Administrator will generate code:

The code generated for a Job only creates the data stores that are specified as output tables for the Job. It does not create every data store in the Process Flow in the right panel of the Process Editor.
The output tables for a Job are listed under the Job's icon in the left panel of the Process Editor. Other data stores in the Process Flow are inputs to the Job.

For example, in Job with Multiple Output Tables in the Process Editor, the code generated for the All Sales Detail Job will only create Sales Detail Grp, Customer, Drop, and so on. It will not create the ODDs in the Process Flow (icons at the bottom of the Process Flows in the right panel of the Process Editor). The ODDs are inputs to the All Sales Detail Job. They must be created in one or more separate Jobs. After the ODD Jobs have been executed, the ODDs can supply information to the output tables in the All Sales Detail Job.

Creating Multiple Output Tables with One Job

You do not have to create a Job for each data store in your project. You can create a single Job that creates multiple data stores.

For example, Detail Logical Table with Detail Tables in the Explorer shows a Detail Logical table (Sales Detail Grp) with Detail Tables (Customer, Drop, and so on). You could define a series of individual Jobs, such as the Customer Job shown in Job and Process Flow in the Process Editor, which would create each of the Detail Tables under Sales Detail Grp.

Alternatively, you could create a single Job that would create Sales Detail Grp and all of its Detail Tables, as shown in Job with Multiple Output Tables in the Process Editor.

Job with Multiple Output Tables in the Process Editor

[Job with Multiple Output Tables in the Process Editor]

To create all of the output tables listed under the All Sales Detail Job (item with the rectangle around it in the left panel of the Process Editor), you would click the Job with the left mouse button and select Run from the pop-up menu. The Load Generation/Execution Properties window would display. From the Load Generation/Execution Properties window, you could execute the code associated with the All Sales Detail Job.

For more details, see Example: Defining a Job with Multiple Output Tables and Input Sources in a Process Flow.

Restrictions on Jobs with Multiple Output Tables

SAS/Warehouse Administrator does not allow you to add an output table that is invalid for a particular Job. All output tables must be in the same metadata repository as the Job.

For example, you can create a Job whose output tables are all in the same Data Warehouse, and you can create another Job whose output tables are all ODDs in the same Warehouse Environment. But, you cannot create a Job in which both a warehouse data store and its ODD are specified as outputs.

For example, if you had to create a set of Jobs to manage the sources and targets shown in Job with Multiple Output Tables in the Process Editor, you could

Create and run one Job called All Sales ODDs that creates and loads all of the toy sales ODDs in the Toy Store Environment.
Create and run another Job such as the All Sales Detail Job shown in Job with Multiple Output Tables in the Process Editor.

For details about the impact of metadata repositories on some user operations, see Overview: Metadata Repositories.

Top of Page