Creating Job Statistics for the SAS Data Integration Studio Job Statistics Reporting Package :: SAS(R) 9.3 Intelligence Platform: System Administration Guide, Second Edition

Summary of Steps for Creating Job Statistics

To create job statistics for use in SAS Data Integration Studio job statistics reports, follow these steps:

Run deployed jobs on the SAS DATA step batch server.
Parse the batch server logs to create job statistics data sets.
Run the second parsing program to create job statistics summary data sets.
Build the job statistics cubes. This step is necessary only if you will be using SAS Web Report Studio to generate the job statistics reports.

You might want to use a scheduling facility system to run the parsing programs and build the cubes on a regular basis (for example, nightly). Doing so will ensure that up-to-date reports are always available.

Details for each step are provided in the following topics.

Step 1: Run Deployed Jobs on the SAS DATA Step Batch Server

The job statistics reporting package collects statistics only for jobs that have been deployed on the SAS DATA step batch server where the reporting package has been installed. Be sure to deploy your SAS Data Integration Studio jobs on this server. See “Deploying Jobs” in the SAS Data Integration Studio: User's Guide.

Step 2: Parse the Batch Server Logs to Create Job Statistics Data Sets

After the batch server log directory has been populated by logs, you can execute the log parsing program. This program extracts information from the batch server logs to create job statistics data sets. The extracted information includes ARM information, error and warning messages, source and target table names, and SAS metadata user identities. Follow these steps to run the parsing program:

On the batch server host machine, go to the subdirectory that contains the batch scripts for SAS Data Integration Studio job statistics reporting (for example, SASDIStudioJobReports/batch). This directory was set up in Step 1: Create Directories on the Batch Server Host Machine.
Execute the following script:
- On Windows: DIStudioJobStats.bat
- On UNIX: DIStudioJobStats.sh
When the program is finished executing, do the following to make sure that it completed successfully:
- Open the library location for the job statistics data sets (for example, C:\SASDIStudioJobStats) and make sure that the following data sets have been created: dijobstepmesg.sas7bdat, dijobstepstatspre.sas7bdat, dijobstepstatspre.sas7bndx, dijobsteptable.sas7bdat, and logfilehistory.sas7bdat.
- Examine DIStudioJobStats.log, which is located in SASDIStudioJobReports/batch/logs.
After you run the parsing program for the first time, update the file DIStudioJobStats.Properties (located in SASDIStudioJobReports/batch) as follows:
1. Find the line that contains DIStudioJobStats.overwriteTables=. Change the value of this property to false.
2. Save the file.

Step 3: Run the Second Parsing Program to Create Summary Data Sets

The second parsing program creates summary data sets from the detailed data. To run this program, follow these steps on the batch server host machine:

Log on as the user who is specified in the DIStudioJobStats.user parameter of the DIStudioJobStats.Properties (for example, sassrv).
From the subdirectory that contains the batch scripts (for example, SASDIStudioJobReports/batch), execute the following script:
- On Windows: joblogsFlow.bat
- On UNIX: joblogsFlow.sh
When the program is finished executing, do the following to make sure that it completed successfully:
- Make sure that the following data sets have been created in the library location for the job statistics data sets: dijobmesgs.sas7bdat, dijobstepstats.sas7bdat, dijobstepstats.sas7bndx, dijobsummary.sas7bat, djobsummarylist.sas7bdat, and dijobtables.sas7bdat.
- Examine joblogsFlow.log, which is located in SASDIStudioJobReports/batch/logs.

Step 4: Build the Job Statistics Cubes for SAS Web Report Studio Reports

The next step is to access the summary data sets and build the physical cubes that are used to generate job statistics reports.

Note: This step is necessary only if you will be using SAS Web Report Studio to generate the job statistics reports.

The user who builds the cubes must have the appropriate permissions as specified in Step 8: Apply Required Permissions.

To build the physical cubes, follow these steps:

Log on to SAS Data Integration Studio. On the Inventory tab, expand the Cube node.
Right-click the DI Job Step Stats cube, and select Build Physical Files.
Right-click the DI Job Step Messages cube, and select Build Physical Files
The SAS log from each job is displayed. Examine the logs to ensure that the jobs ran successfully.
Note: When you build the physical cube for DI Job Step Messages, the following errors are displayed if your batch server log files do not contain any warning or error messages:
```
ERROR: The input data set does not contain any observations.
```
```
ERROR: The input data set cannot be opened.
```
```
ERROR: Cube "DI Job Step Messages" cannot be created.
```
If these messages are displayed, the job statistics reports that analyze error and warning messages will be unavailable. You do not need to take any actions to resolve this issue.

After the cubes are created, you can access the SAS Data Integration Studio job statistics reports.