Did something on n8n’s side change? All of our automations began failing around 12:40 pm without any changes made on our side
We are running version 1.36.1 on n8n cloud.
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:
version 1.36.1
database, executions_process setting - ?
n8n cloud
Windows 11
What are those workflows?
How did they fail? On which steps? Are they all failing on the same spot?
This would most likely be one service you’re using being down or credentials expiring.
There is not much we can see from just this screenshot. Please add more context if you’re able
Hi Liam, thanks for your message. They appear to be failing on the same GraphQL Authentication step with the message "Response body is not valid JSON. Change “Response Format” to “String”.
These workflows have been running fine for months without this error.
Try setting up the auth again and getting it connected manually and see if the issue persists.
If not please send the exact error messages as well as input data so we can try to recreate
Hi Liam - that didn’t work. We have the same GraphQL node in-use for 4 other endpoints and the only difference between them is the endpoint - even login details are the same. Only one of the endpoints is suddenly experiencing this failure while the others are still outputting a valid JSON response body.
Converting the output to string from the node with a failing endpoint outputs a text with <!DOCTYPE html><html lang="en-US"><head><title>Just a moment...</title><meta http-equiv="Content-Type" content="text/html; charset=UTF-8"><meta http-equiv="X-UA-Compatible" content="IE=Edge"><meta name="robots" content="noindex,nofollow"><meta name="viewport" content="width=device-width,initial-scale=1"><style>*{box-sizing:border-box;margin:0;padding:0}html{line-height:1.15;-webkit-text-size- …
More details for the error:
Stack trace
NodeOperationError: Response body is not valid JSON. Change "Response Format" to "String" at Object.execute (/usr/local/lib/node_modules/n8n/node_modules/n8n-nodes-base/dist/nodes/GraphQL/GraphQL.node.js:446:35) at processTicksAndRejections (node:internal/process/task_queues:95:5) at Workflow.runNode (/usr/local/lib/node_modules/n8n/node_modules/n8n-workflow/dist/Workflow.js:728:19) at /usr/local/lib/node_modules/n8n/node_modules/n8n-core/dist/WorkflowExecute.js:660:53 at /usr/local/lib/node_modules/n8n/node_modules/n8n-core/dist/WorkflowExecute.js:1062:20
Expected JSON output working in other nodes:
[
{
"data": {
"login": {
"accessToken": {
"token": "(token value)"
},
"refreshToken": {
"token": "(token value)"
}
}
}
}
]
Thanks for the extra info.
I won’t be able to look at this now so I will send to the n8n team.
That might be better anyway since it is cloud
good luck
1 Like
@netroy Are you available to help?
Hey @pmg,
Can you share the full output you are getting it looks like html so it will likely show what the issue is.
1 Like
Hi Jon, I keep getting a 403 error when trying to share the full string html output here unfortunately.
Here is an example - this endpoint appears to be failing in all nodes where it is used as of a day ago. We use the same authentication nodes for many different workflows. Any nodes using a different endpoint are still fine while any nodes using this endpoint in the failing node are failing in all workflows using it.
One of the last times the (now-failing) nodes with an error were working at 12:36:37 on April 22 - before all started failing at 12:40.
Hey @pmg,
Can you save it to a file and share that file? Have you tried the same request from Postman or similar as well to make sure nothing has changed with the API as well like maybe something like Cloudflare being introduced?
Actually this is exactly the problem, Loading https://api.pets4homes.co.uk/api in my browser is file and from n8n and curl it fails with a similar looking html page but as soon as I add a user agent header it works.
I suspect they have recently updated their Cloudflare rules to block certain user agents or agents with only a few characters in the name…
We send axios/1.6.7 as the user-agent which fails…
By changing it to n8n it works
1 Like
@Jon Fantastic - that seemed to work! I’ll update this everywhere on our end. Thanks for being a lifesaver.
1 Like
Hey @pmg,
No problem, For a minute you had me worried that we had broken something 
This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.