Tracking Runs and Errors
With workflow created, you need to know how to track progress and deal with potential errors
The goals for the article are:
- Understand the parts of a Workflow Run
- Be able to read and respond to Errors
- Know how to Test and Rerun workflow
A Run represents one time through the workflow process (tree). Depending on the tree configuration (connector qualifications and input data for example), each run could be very different.
Here is an example of a "basic" run:
The default view for a Run is the Timeline. It includes all the elements of a Run (see above image).
Other information available from this view include:
- Run ID - each run is identified by a specific number
- Source|Group|Tree|ID - identifies the tree that was used OR
- Global Routine - identifies the routine that was used. Routines include a link to the Parent Run
- Refresh - it can take sometime for a tree to process
- Actions - two actions are available from the run console. Create Manual Trigger lets you start a tree from a specific node if it has an error or otherwise stalls. Run Again lets you re-run the tree with the same inputs. Extraordinarily helpful for testing or troubleshooting (edit inputs!).
- Builder - Opens the builder for the associated tree
Below all of these is a table of all the elements for the tree. Select one of the elements to only display those elements.
Nodes and Results
There are two main elements that make up a run, Tasks and Triggers. Each task entry represents a node being processed. Depending on what type of node it is, you may have results that can be viewed in the run.
Example Task with Results:
Editing Results is helpful if you are creating a manual trigger and you want different data processed.
Triggers are only created when the engine needs to start a process or perhaps restart a deferred node. There is always a Start Trigger, and some utilities need triggers like Loops.
The inputs shown in the Run are the values that are used for that execution of the tree. The format and structure of the input depends on the Source. For Request, the input is a JSON string that is used to populate the Parameters for a node.
Example of Request Source:
The inputs can be edited. This is helpful if you are troubleshooting an error or testing new workflow.
When an error happens in the workflow it is reported through the Run.
If you can resolve the error, the error console provides a great option for continuing the workflow process.
- Do Nothing - resolves the error and stops processing
- Retry Task - will retry the node with the existing parameters
- Skip Task - skips to the next node and continues the process
For each action, there is a required Notes field. This information is stored with the error.
In the Next Article
In the next article, we'll go over the included System Utilites options.
Updated over 1 year ago