Troubleshooting

To troubleshoot pipelines created in the IIP, the first step is to identify which pipeline encountered an issue. The next is to find the Snap error messages for debugging.

Find the problem pipeline

Finding the pipeline to debug can be challenging, especially in environments with tens of thousands of pipelines in production. The error report might come from an external source such as Zendesk. External incident reports often refer to the data by some unique identifier, such as a purchase order number. Because SnapLogic does not store record data or metadata, a simple search in the Execution overviw cannot identify the pipeline that processed those records.

Encourage pipeline developers to use the best practices that enable others to find pipelines:

  • Name pipelines using a convention that helps others understand their purpose.
  • Organize related pipelines in logically-named projects in the same Project space. The Execution overview search provides Project Space and project search under advanced options:
    Project space search

  • Document the pipeline purpose and the systems it interacts with. This might be in a document such as a spreadsheet.
  • Use pipeline parameters to capture a value that maps to a record. The Execution overview string search returns results for parameter values.
  • Use error pipelines to output and store record data when a failure occurs.

Identify the Snap error

After locating the problem pipeline in the Execution overview, identify which Snap or Snaps encountered an issue. Click the pipeline to open the details pane. From the details pane, try the following to learn more about the error:

  • Use the Error filter on the Snap statistics and Pipeline logs tabs to bring errors to the top:
    Errors filter

  • Use the All filter on the Pipeline logs tab to find out what happened just before the error.

In the case of errors caused by the data itself, such as formatting issues, you will need to find the issue and fix it in the source. This is easier for pipelines with an error pipeline that stores record data when a failure occurs.