Forbidden - perhaps check your credentials after restart on self-hosted

Questions
1

Describe the problem/error/question

We have a self-hosted n8n instance with several active workflows. One of our workflows uses an AWS SNS Trigger. Each time we restart n8n (usually after upgrades), this particular workflow fails to reactivate, showing an error. Deactivating and immediately reactivating the workflow resolves the issue, but this creates a manual step in our upgrade process that we’d like to avoid.

Is there a way to prevent this from happening, or is this a known issue with the AWS SNS Trigger?

What is the error message (if any)?

Problem activating workflow
The following error occurred on workflow activation:
Forbidden - perhaps check your credentials?

Please share your workflow

Not shared due to proprietary information

Share the output returned by the last node

The workflow does not start or receive any triggers until manually reactivated. No output is produced until this step is taken.

Information on your n8n setup

  • n8n version: 1.101.1
  • Database (default: SQLite): Postgres (hosted on AWS)
  • n8n EXECUTIONS_PROCESS setting (default: own, main): (default)
  • Running n8n via (Docker, npm, n8n cloud, desktop app): Docker (cloud container host)
  • Operating system: Ubuntu (cloud)

Additional Details

  • I found a similar issue on the forums, where the recommended fix is to manually reactivate the workflow.
  • We upgrade n8n regularly, so having to manually reactivate this workflow each time is not ideal.
  • The AWS credentials work fine once the workflow is manually reactivated.
  • This only affects the SNS Trigger workflow—other workflows do not have this issue.

Is this a bug with the AWS SNS Trigger, or is there a recommended workaround for persistent activation across restarts?

2

This is a common issue even when I use any node of google like sheets or drive, it automatically signs me out after 2-3 days and then I have go again in credentials part of those nodes to sign-in again. So this is a problem from the local hosting part created by n8n for which they might give solution in future.

3

Interesting. In my case I don’t have to do anything with the credentials for it to work again. All I do is deactivate the workflow and reactivate it and it starts functioning normally again.