The referrer does not match the host.

[vinooganesh]

Select Topic Area

Question

Body

Hi All, I'm attempting to run an installed airflow in Github codespaces, but when I open the port that the webserver is running on and try and execute anything - I get a "400 Bad Request - The referrer does not match the host." I'm not familiar with how the network setup of codespaces works, but I'm concerned that nginx is stripping out the headers. Can someone help me confirm what's going on?

Repro steps:

1. Run pip install "apache-airflow==2.5.1" --constraint "https://raw.githubusercontent.com/apache/airflow/constraints-2.5.1/constraints-3.7.txt"
2. Run airflow standalone
3. Attempt to open port 8080 on the box and log in with the admin password that is output at the end of the airflow standalone command
4. You'll hit this error.

[jge162]

Hello and good day @vinooganesh this error occurs when the webserver is configured to check for a specific value in the -> referrer header, and the value being sent in the request -> does not match.

• Check your configs and/or ->
• try setting referrer header in your request manually to see if that resolves the issue.

[vinooganesh]

Hey @jge162 thanks for the quick response! The challenge is that I can't easily control the headers given that this is supposed to be open-source/standalone airflow. Meaning it's the actual airflow webserver talking to the workers (outside of my control). Is it possible to make sure the nginx config isn't dropping these headers?

[Jainil-007]

Hi @vinooganesh, I am following your "Hands-On Introduction: Data Engineering" and stuck on the same issue, it would we great if you can explain the workaround in a bit detail so I can continue with the course :)

[vinooganesh]

Hi @Jainil-007 - the workaround is actually built into the steps in the course. It involves really just refreshing the page and changing the URL (which was intended just as a workaround for airflow, not a generic solution).

[zselyigy]

The method suggested in the course video worked for me yesterday (the error page was similar what is in the video), but not working today (the error page is different "Bad Request The referrer does not match the host.")

I tried to modify the url according to that can be seen in the video (adding preview. in and changing /login to /home at the end of the url) and it led to another error: {"type":"https://tools.ietf.org/html/rfc7231#section-6.5.4","title":"Not Found","status":404,"traceId":"00-d2d1fa4ac623e6f43ec85e29c5adc27e-11a5b99ea6432298-00"}

Please help us to fix it!

[vinooganesh]

Hi @zselyigy, I spent some time looking into this yesterday for the LinkedIn Learning airflow course specifically. I'm going to look into a longer term workaround since it seems to be hitting a few people. There is likely some kind of race condition in updating the WTF_CSRF_ENABLED setting with this command: sed -i -e '/WTF_CSRF_ENABLED =/ s/= .*/= False/' ${AIRFLOW_HOME}/webserver_config.py. It seems like sometimes if airflow takes too long to create the webserver_config.py file, then the sed command runs first and fails, and airflow's webserver instead starts up with the default setting for WTF_CSRF_ENABLED which is True. That results in the "Bad Request: The referrer does not match the host error"

There are a few potential fixes, but the course is actually likely going to be reworked a bit as Airflow itself has updated, so some of the instructions are different in more recent versions of Airflow.

To fix this Bad Request issue for now, can you make sure that WTF_CSRF_ENABLED in webserver_config.py is set to False? You'll need to restart the airflow webserver for this to take effect.

I'm going to work with the LinkedIn Learning team on making this process a bit soother - we may even move to airflow standalone or running airflow inside of a docker container on the machine.

cc @Jainil-007 @frans-k

[docmhvr]

Hello,
Making CSRF as False and restarting the webserver and scheduler works!
It first gives error not found on refreshing however going back and logging in again gets the servers and DAGS up and running. I can see the Airflow UI and created DAGS!
Thank you for this discussion helped solve the error.

[github-actions[bot]]

🕒 Discussion Activity Reminder 🕒

This Discussion has been labeled as dormant by an automated system for having no activity in the last 60 days. Please consider one the following actions:

1️⃣ Close as Out of Date: If the topic is no longer relevant, close the Discussion as out of date at the bottom of the page.

2️⃣ Provide More Information: Share additional details or context — or let the community know if you've found a solution on your own.

3️⃣ Mark a Reply as Answer: If your question has been answered by a reply, mark the most helpful reply as the solution.

Note: This dormant notification will only apply to Discussions with the Question label. To learn more, see our recent announcement.

Thank you for helping bring this Discussion to a resolution! 💬