The exception task writes the results of an exception task job to the exception data store or to a comma-separated file in a directory that the Secure Agent specifies in your local system. You can download the records that the task identifies as exceptions from the exception data store.
To retrieve the job results data from the exception data store, check the session log for the job in which the task ran. To retrieve the job results data from local storage, browse to the directory that the flat-file connection specifies in your local system. Find the directory details on the Connections tab in Administrator.
You can analyze or resolve the data quality issues that the exception records contain in a separate data process. If your organization no longer needs the exception records, you can delete the records and the associated exception indicators.
If you no longer need an exception task, you can delete the task from the Explore page. When you delete an exception task, the operation also deletes the exception records and exception indicators for any job that ran for the exception task from the exception data store.
To download exception records and delete exception data, your user role must have the Exceptions Data - View and Exceptions Data - Delete feature privileges. Additionally, you must have Read and Execute permissions on an asset to download the asset, and you must have Delete permission on an asset to delete the asset.
Downloading and deleting exception data
Download and delete exception data from the exception data store on the My Jobs page.
You can open the page in Data Quality, Data Profiling, or Data Integration. You can also view the details of the exception task job on the My Jobs page.
1Open the My Jobs page. The page lists all of the jobs that you run.
2On the My Jobs page, click an exception job name. The page displays detailed information about the exception job.
The following image shows the details of an exception job:
The job details include the following statistics:
- Success Rows. The total number of rows that the job read.
- Exception Rows. The number of rows that contain data quality issues.
- Number of Issues. The total number of data quality issues that the job identified across the rows that it read.
3In the Results area on the exception job page, you can perform the following operations:
- To download the exception data associated with the exception job, click Download Exception Data.
You download the data in a comma-separated format.
- To delete the exception data that the job identifies, click Delete Exception Data.
- To download the session log for the exception task job, click Download Session Log.
The Download Exception Data and Delete Exception Data options appear disabled for a job instance in the following cases:
- You deleted the exception data for the job instance.
- You deleted the exception task that generated the data.
You can stop and resume a job on the My Jobs page.
Rules and guidelines for exception data
Informatica Intelligent Cloud Services uses a range of properties to manage the exception data that your exception jobs generate.
Consider the following rules and guidelines when you download and review exception job data:
•The exception data store can store up to 20,000 issues for a given exception task job.
•An exception data file that you download for a job from the exception data store can contain up to 20,000 issues.
•By default, the exception data store retains exception job data for 45 days. Informatica Intelligent Cloud Services deletes data for jobs outside the 45-day time frame. To request a shorter default time frame for your organization, contact Informatica Global Customer Support.
•When you opt to store job data on your local system, the exception task writes the results as a comma-separated file to the flat file connection that the Secure Agent can access. Informatica does not impose a limit on the quantity of data that the exception task writes to the file.
•You can set up permissions to protect your data from other users. For information about privileges and permissions on the exception tasks, see Exception management prerequisites.
•The application in which you open the exception data file may apply a limit to the quantity of data that it displays. For example, Microsoft Excel places a limit on the number of rows and columns that it displays in a single worksheet. Consider creating a backup copy of the file that you download before you open and save in another application.
Reading the exception task job data
The exception data that an exception task job generates contains columns from the source data set that the task reads and columns that the exception task creates at run time. The columns that the job adds provide additional information about the task and how the job identified the exceptions.
The exception task job creates the the following columns:
Row ID
Indicates the row number of the exception record in the source file.
A given record can appear more than once in an exception report in the following cases:
- A rule specification found a data quality issue in more than one input column in the record.
- More than one rule specification identified a data quality issue in the record.
When more than one instance of a record appears in the exception report, each instance has the same Row ID value.
Priority
Indicates the priority of the data quality issue that defines the record as an exception. The exception task reads the priority value from the rule specification that identifies the record as an exception.
Exception Message
Describes the data quality issue that defines the record as an exception. The exception task reads the exception message from the rule specification that identifies the record as an exception.
Rule Name
Contains the name of the rule specification that identifies the record as an exception.
Rule Input Columns
Identifies the columns of input data that the rule specification in the exception task analyzes.
Job ID
Contains a unique identifier for a particular job that you ran.
Creation Date
Contains the date on which you ran the exception job.
Dynamic renaming of duplicate field names
If the input data to the exception task contains a column with same name as a column that the job adds to the exception data, the system renames the input column in the following ways:
•For the column names Row ID, Exception Message, Rule Name, Rule Input Columns, and Creation Date, the system replaces the space character in the input column name with an underscore. For example, the system changes the name Rule Input Columns to Rule_Input_Columns.
•For the column names Job ID and Priority, the system appends an integer to the input column name. For example, the system changes the name JobID to JobID_1.
