Retry failed execution with currently saved workflow

[jb8n]

Hi,

I just went crazy after an hour of debugging n8n.

It appears there’s a bug in the feature “Retry with currently saved node”, as reported last year, now closed for additional comments:

Bug: Retrying a failed execution with the currently saved workflow still executes old node state

Describe the problem/error/question I have a failed execution for a node containing a code node. I was changing the code in the code node but when I retried the execution with the currently saved workflow, I still got the same error which cannot be as I changed the code node entirely. So, I duplicated the code node. I unwired the original code node and wired up the copied code node in the workflow. When I retried the execution again, still the new code node was not executed but the original one…

Edit:
I think I reported the same bug, which actually isn’t a bug, but the end-user not reading properly

My simple workflow is: Webhook → Node code → Telegram node

The node code does sanitization and string transformation.

I had an issue in the Node code, making the Telegram node fail. Therefore I iterated in the node code.

However, the retry feature doesn’t retry with any of my subsequent changes. The execution view UI explicitly mentioning “currently saved workflow (from node with error)”

In my case, that feature doesn’t add value because the error that made the whole execution fail, is after the node that I am debugging and iterating on.

My question is then, how can I retry a whole workflow after editing parts of it, prior to the node that failed?

• Is there a “Retry with currently saved workflow (entire wokflow)” ?
• From the failed execution, when I click “Debug in Editor”, it doesn’t work for me as it doesn’t pin the webhook payload.

Setup

Selh-hosted LAMP server with n8n Docker 1.94.0

[Umair_Khan44]

are you using self-hosted if you are then you should use tunneling because localhost doesnot work
and you donot use tunneling in cloud

[nadine]

Thanks for picking up on this @jb8n . I totally agree that there should be a “Retry with currently saved workflow (entire workflow)” option as in some cases “Retry with currently saved workflow (from node with error)“ doesn’t make sense (because you want to fix the workflow in a way that prevents the error based on previous nodes) Can someone from the team pick this up? Perhaps @ria ? BTW, this is not related to a specific workflow. You can easily reproduce this retry behavior with any workflow.