We’re excited to roll out a much-requested improvement to logging today! Starting today, you’ll notice a new type of server logs in your logs tab— “Error running API workflow”. Here you will find logs flagging when an API workflow errors out. Previously, when an API workflow failed, we wouldn’t log anything, making debugging difficult.
Astute observers might notice that the ID of the specific query or action that fails is also now logged in the UI. Advanced users of dev tools can use this to quickly identify the exact step of the workflow that failed. In the future, we plan to flesh out this UI to allow you to easily jump to the particular action / query that failed without needing to use dev tools.
FAQs
What’s the difference between “Workflow error” and “Error running API workflow”?
“Workflow error” contains errors generated by non-API workflows, whereas “Error running API workflow” contains errors generated by API workflows.
Great news! Next can you please allow exporting of these logs? The default view in bubble is super inefficient so it’s hard to to go through. When I have to search through logs I just cringe…
Would this allow us to see which condition made the workflow or actions of the workflow not run ? I often have multiple conditions on an API workflow and if I get Action Conditions failed I know I am f…… Also because there is no easy way to remove conditions in the middle of the expression without having to rewrite the next part
@henry.dowling Please add a feature for front end and backend workflows to have a trigger based Error running API workflow just like there is on the front end for general workflow error.
If we have this feature of a trigger when there is an error running API workflow we can create systems that help us mitigate the issues for our users when one of these errors occur.
Having the logs show us this information is a great step forward, but we need the feature to trigger actions based on when the API workflow has an error, because currently when an API timeout occurs all that Bubble does is terminates the workflow, which basically makes it impossible for us to develop ways to mitigate issues when they API timeout occurs.