3.2. Monitoring & Debugging Workflows

Modified on Wed, 12 Nov at 6:05 PM

Monitoring and debugging a workflow involves tracking its execution to ensure it runs smoothly and identifying any issues that may arise. Monitoring enables real-time performance checks, while debugging involves analyzing logs and error messages to identify and troubleshoot problems.

In this section, we will learn how to monitor and debug a TeamPal workflow.

1. Monitoring

After creating a workflow, you will need to test it to ensure it functions correctly. You can do that using the “Test workflow” feature.

The dashboard will display the currently running node, along with the previous nodes, in real-time. You can review the overall flow and verify your logic if something seems incorrect.

After executing the workflow, the credits used will be displayed and charged. For information about our charging method, please visit the “Workflow Credit Explanation” section. Additionally, the workflow history provides an overview of the automation history, allowing you to differentiate between runs based on their execution times.

2. Debugging

When running the workflow, you may occasionally encounter a logic error, which can lead to an error node. In such cases, the workflow will stop, and the error node along with the error message will be displayed. Below are some common errors you might encounter during this process.

Node	Error/Obstacle	What needs to be checked	Note
(Blank)	Test workflow buttons/Automation settings were disabled	The edges and parameters in your workflow	The buttons will be disabled unless there is a feasible path (with an edge connected) that connects Begin to End
Python	Error: Function execution failed, please check the code of the function. Detail: SyntaxError: closing…	Check the syntax of your Python code
Python	Error Error: Function execution failed, please check the code of the function. Detail: TypeError: can only…	Check the input parameter types of your Python code	Maybe there is a mismatch between the parameter type of the previous node and the parameter type declared
Knowledge retriever/LLM	Error sequence item 1: expected str instance, int found	Mismatch parameter type	Input into the parameter of the knowledge retriever/LLM must be in “string” type
Knowledge retriever	Blank item when passing the result to another node	Your [n] input must be out of indices