Http request doesn't work on n8n but on make.com and browser

Questions
1

Using this http-request:

https://browser.ihtsdotools.org/fhir/ValueSet/$expand?url=http://snomed.info/sct?fhir_vs&count=1&offset=1&filter=pain

works in postman and in the browser as well as in make.com but not in the n8n http-request-node.

The circle in the Execution-window rotates after testing this step …no answer
Any ideas? I tried all day long…asking claud, open.ai…no solution for me…

Running n8n via n8n cloud
Operating system: macOS

This is my workflow:

2

It looks like your topic is missing some important information. Could you provide the following if applicable.

  • n8n version:
  • Database (default: SQLite):
  • n8n EXECUTIONS_PROCESS setting (default: own, main):
  • Running n8n via (Docker, npm, n8n cloud, desktop app):
  • Operating system:
3

@dranbecker Not working for me neither. Works in the browser and in html and raw json. Not sure why it does not return a result in n8n. It does say: The service refused the connection.

4

Thank you… i used the devTool in chrome, copied header exactly…doesn’t work…

5

From what I can tell, the service has some anti-scraping mechanism in place. I got a couple of 429 Too Many Requests, and then my public IP was immediately blocked.

To me it sounds like this service doesn’t want to used via 3rd parties like this.
So, even though it might be possible to fix this in n8n, I’m not sure if it’s ethical for us to do so.

If you can share public documentation for this service that states how to use this service, then we can try to get this fixed.

6

Snowstorm FHIR 4.0 provides access to SNOMED CT. This is an international classification of medical terminology that I use in an N8n workflow to extract symptoms with the help of an AI agent, and then forward those symptoms to a publicly accessible API. As described in my first post, it works smoothly with make.com and Postman. Unfortunately, as you’ve already seen, it does not work with N8n’s HTTP Request node.

I’ve already considered a workaround, but it would be quite complex—storing the data somewhere else in a different database first, querying that data with another app like make.com, and then bringing it back into N8n. I’d like to avoid that if possible. I’ll attach a link with documentation for the relevant requests. In any case, thank you very much for your help.

Link to snomed

7

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.