Extracts Process Diagnostics Report

You can run the Extracts Process Diagnostics Report process to generate the diagnostics report of an HCM Extracts definition or for a specific HCM Extracts process.

This process generates the Extract design time, runtime, data, and changes only extract diagnostics report to:

View recommendations about an Extract definition design, missing or incorrect design choices, and take corrective actions.
Identify the phase-wise time and recommendations about an Extract run to improve the processing time for the next runs.
Identify the anomalies in the extracted data that prevent the usage of generated HCM Extracts XML data by Oracle Business Intelligence Publisher or other publishing tools.
Debug incremental extract changes with the Changes Only modes.

To submit and run the process from the Scheduled Processes work area:

Click Navigator > Tools > Scheduled Processes.
On the Scheduled Processes Overview page, click Schedule New Process.
Leave the type as Job, search and select the Extracts Process Diagnostics Report process, and click OK.
In the Process Details dialog box, select the Scope and Scope Value:
- For Design Time Diagnostics scope, the available extract definitions are listed as the scope value.
- For Runtime Diagnostics, Data Diagnostics scope, and Changes Only Diagnostics the extract runs or instances are listed as the scope value.
For Changes Only Diagnostics, select the mode category and attribute name and ID.
Define the schedule, output, and notifications for the process set, as you’d do for any scheduled process. Set any other options and click Submit.

The runtime and data diagnostics can be generated for any extract instance, whereas the design time diagnostics is generated for user-defined extract definitions. After the process is complete and successful, review the log and the process report to know about dependencies, warnings, and potential solutions.

Design Time Diagnostics

The Design Time Diagnostics for a user-defined extract definition provides the following information:

How to reset the context user entity
Tips on using threading database item usage for incremental and changes-only extracts
Considerations for historic user entities in threading data groups
Delivery options and the associated BIP reports, if any
Design practices that are commonly seen to impact extracts performance including the usage of advanced conditions, override conditions, and record the sort order

These recommendations can help improve and optimize the Extracts design.

Note: You can also run the process for design time diagnostics from the Data Exchange work area> Extracts Definitions page. Search for and select an extract definition and click the Validate button. In the Execution page, click the Validate > Run Extended Validation. The action automatically triggers the process. You can download the results and view the log details.

Runtime Diagnostics

The Runtime Diagnostics for a specific HCM extract provides the following information:

Parameter values for the extracts run
Volume of data processed by the extracts run
Delivery status for the delivery options defined in the extract, including the BIP job IDs and job statuses
Phase-wise timing of the extracts run includes the start and end time of each of these phases: Archive, XML generation, BIP job process times, and Delivery. This provides an understanding of performance issue and suggests resolution too.
Tips to improve the process timings. You might want to consider them when updating or improving the extracts design.

Data Diagnostics

The Data Diagnostics for a specific HCM extract data provides the following transformation and parsing information:

Transformation details such as:
- Instance name and library used
- Extract Definition ID and name
- XML tag name of main block and database item (DBI)
- Number of processed and error records
- Status of the transformation, with start and completion time
Problem descriptions such as:
- Invalid characters in the output
- Empty DBI tag name, without threading DBI
- Error details listing the Object Action ID number, its location, and tag name

Changes Only Extracts Diagnostics

The Changes Only modes in the diagnostics report are used to debug incremental extract changes to generate the current status of the data, compare it with the baseline data from previous runs, and identify anything new and modifications. The modes are:

Previous Extract Run: It is used for basic Changes Only extracts with parameter CHANGES_ONLY = Y
Object Extracted But Not Expected: It is used for advanced Changes Only extracts with parameter CHANGES_ONLY= ATTRIBUTE, ATTRIB_OLD, BLOCK_OLD, or BLOCK
Object Not Extracted: It is used for data that was extracted in the past runs but isn't extracted in the current run. You've to provide the threading object ID of the missing object from the earlier run to identify the changes.
Record Not Extracted: It is used when the data record for the marked primary key isn't extracted in the current run. To identify the changes, you've to provide the Threading Object ID, select the extract record name, and enter the pipe delimited name value pair (for example: Extract All Phone Number=6623889833||Extract All Phone ID=300100625351893||) as key attributes that weren't extracted from the earlier run.
Unwanted Record Extracted: It is used when an unintended record and data elements are extracted to find the intended changes in the record attributes. Based on the attribute changes, the report displays the record status as changed, added, or deleted. To identify the changes in the record attributes, select the Object Action attribute and provide the Object Action ID, select the record name and enter the name value pair (for example: Extract All Phone Number=6623889833||Extract All Phone ID=300100625351893||) as key attributes to find the threading data element and primary key, and do the comparison between the past and current run.