Inspect history and retry execution

Inspecting the history is useful to understand the behavior of the integration for specific messages and amend functionality or incoming messages.

On the integration click the tab History to see a list of all historic, and current, executions, and their status.

Click on the id of a specific execution to see details about the execution in a new browser tab.

You can review the step execution sequence of the machine workflow in a list. You can also see the execution visualized in view similar to how the workflow is presented in Flow 6 Studio.

If you have turned on "Enable inspecting input to integration" and/or "Enable inspecting output of integration" in the Diagnostic settings you will be able to inspect the Input and/or Output to and from the integration execution instance.

If you have turned on "Track variable changes" in the Diagnostic settings you will be able to inspect inspect individual steps of the executed machine workflow by clicking the link of the individual step.

In this view you see the entire variable set of the machine workflow before and and after the step.

You can also toggle Diff mode to On, and by doing so you can compare the variable set before and after the step side by side.

If you are inspecting an integration that has failed you will notice the button Retry in the upper right of the window is now possible to click.

This is useful in the case the cause of the error appearing in your workflow execution can been fixed outside of the integration and the workflow. This can be the case for example if the connector of a machine step is misconfigured or for some other connectivity issue cannot execute. Such an error cause can be remedied for example by updating the connector configuration or re-enabling the connectivity. When the cause of the error has been remedied you can Retry the integration execution by clicking the button. The integration execution will then pick up again from after the last successful step. So no previous successful steps will be re-executed, this is particularly important to know when it comes to machine steps.

Sometimes the cause of the error is not of such a kind that the execution can merely be retried. This can for example happen when the system sending the input message to the integration has sent an erroneous message. In such a case it would be best if the sending system could send a new correct message (according to the principle of correcting as close to the source as possible). But sometimes that is not practically possible, or to cumbersome. In such a case, you have the technical possibility to change the variable set inside the execution history view and the Retry the execution. We stress technical possibility because although it is possible, it is certainly not always the right thing to do from a data integrity, auditability and accountability perspective.

But assuming that you know what you are doing when excercising the possibility then this is how to do it.

Click on the step where the error has occurred (the red row). And then change the JSON structure in the text box Before step. When done click the button Modify variables.

Confirm.

Verify the this information message is in line with what you expect from your changes.

Now click the button Retry in the upper right of the window.

Confirm.

The execution instance now executes again and enters Pending state.

Hopefully this fixed the cause of the error and the execution now completes successfully.