Failure Analysis

Failure analysis is a central failure mechanism. Here, the user can identify the failure reason. Failures of any pipeline stored at a particular location(collection). From there you can query your failed data in the Failure Analysis UI. It displays the failed records along with cause, event time, and pipeline Id.

Check out the below given walk-through for failure analysis in the Pipeline Workflow editor canvas.

• Navigate to the Pipeline Editor page.

• Click the Failure Analysis icon.

• The Failure Analysis page opens.

• Search Component: A Search bar is provided to search all components associated with that pipeline. It helps to find a specific component by inserting the name in the Search Bar.

• Component Panel: It displays all the components associated with that pipeline.

• Filter: By default, the selected component instance Id will be displayed in the filter field. Records will be displayed based on the instanceid of the selected component. It filters the failure data based on the applied filter.

Please Note the Filter Format of some of the field types.

Field Value Type	Filter Format
String	data.data_desc:” "ignition"
Integer	data.data_id:35
Float	data.lng:95.83467601
Boolean	data.isActive:true

• Project: By default, the pipeline_Id and _id are selected from the records. If the user does not want to select and select any field then that field will be set with 0/1 (0 to exclude and 1 to include), displaying the selected column.

Please Note: data.data_id:0, data.data_desc:1

• Sort: By default, records are displayed in descending order based on the “_id” field. Users can change ascending order by choosing Ascending option.

• Limit: By default, 10 records are displayed. Users can modify the records limit according to the requirement. The maximum limit is 1000.

• Find: It filters/sorts/limits the records and projects the fields by clicking on the find button.

• Reset: If the user clicks on the Reset button, then all the fields must be reset with a default value.

Cause: The cause of the failure gets displayed by a click on any failed data.