Error when setting up automation

Hi. Anyone experienced this error? Baserow is self-hosted on AWS ECS Fargate running 2.0.6.

Error is throwed when trying to apply a filter or adjusting any settings on any database specific nodes.

image

{"detail":"Authentication credentials were not provided."}

WARNING 2026-02-12 11:32:57,078 django.request.log_response:241- Unauthorized: /api/automation/node/71/

The automation is able to pull data from the database, so it doesn’t seem to be a real authentication issue. However, I get that error when trying to apply filters.

Hi @patrick,

Could you describe your workflow for me, such as the trigger and the various table-changing nodes you’re implementing? A screenshot of their configuration would be great. What filter(s) are you applying?

Cheers,

Peter Evans

Hi Peter. This seems to happen as soon as I start adjusting the refinements in any node where a database is involved.

Testing the node works, and it’s able to interact with the table chosen, but not in accordance with the refinements added.

Hmm, this is a strange one @patrick!

If you modify records in other modules (e.g. Application builder → can you update elements, Database → can you update fields)? I can’t immediately think of anything automation-specific that would cause this.

Cheers,

Peter Evans

I hadn’t touched application builder in a while, but after checking I do see the same there. I have sent a support request about this.

Good morning @patrick,

Would you mind checking a few things for me:

  1. If your AWS WAF (Web Application Firewall) is enabled, can you check CloudWatch for blocked requests?
  2. Do you have AWS ALB (Application Load Balancer) enabled? Can you check the rules to see if there could be a misconfiguration there?
  3. If you’re using CloudFront, can you look under “Behaviours” → “Allowed HTTP Methods”? There’s a small chance the HTTP methods are limited to just reads (GET/HEAD).

Something else you can try if you use ALB: can you send an API request (e.g. modifying a node, link you did in this topic) directly to the Fargate container (to bypass ALB)? If it works, then we’d know if it’s a problem in AWS’s networking configuration.

Cheers,

Peter Evans

Thanks Peter! It was an AWS WAF managed rule issue. Didn’t even cross my mind. Thank you!

1 Like