Desk fails to load: Refused to execute script/style due to incorrect MIME type

[NeihTzxc]

The web interface fails to load properly (resulting in a blank page or broken UI) because the browser refuses to execute CSS and Javascript files. The console shows a MIME type mismatch, where files expected as text/css or application/javascript are being served as text/html. This usually indicates that Nginx cannot find the physical files and is instead serving a 404 error page in HTML format.

Context information (for bug reports)
Framework: Frappe version-15
Environment: Dockerized deployment using frappe_docker
Build Command:

docker build --no-cache \
  --build-arg=FRAPPE_PATH=https://github.com/frappe/frappe \
  --build-arg=FRAPPE_BRANCH=version-15 \
  --secret=id=apps_json,src=apps.json \
  --tag=image-name:tag \
  --file=images/custom/Containerfile .

Steps to reproduce the issue:

1. Build a custom image using the provided docker build command.
2. Deploy the stack using Docker Compose.
3. Access the Frappe Desk via a web browser.

Observed result:

• Browser Console Errors: Refused to apply style from ... because its MIME type ('text/html') is not a supported stylesheet MIME type.
• Multiple 404 Not Found errors for assets located in /assets/frappe/dist/.
• Uncaught TypeError: Cannot read properties of undefined due to missing JS dependencies.
  Expected result:
• All assets should be served with the correct MIME types (text/css or application/javascript).
• The Frappe Desk should render and function correctly.

[DanielRadlAMR]

Are you using the latest version of this repo — the up-to-date main branch?

[NeihTzxc]

Are you using the latest version of this repo — the up-to-date main branch?

Yes, i try git pull for get latest this repo

[DanielRadlAMR]

i can't reproduce this issue. Should be fixed with #1894

[NeihTzxc]

i can't reproduce this issue. Should be fixed with #1894

I tried again, I removed all my custom apps and replaced them with Frappe LMS, and the error stopped happening. I still don't understand why this error occurred; it might be due to my computer.

[DanielRadlAMR]

The whole point of Docker is to make environments machine-independent, so this issue shouldn’t be caused by your PC.

Did you install the applications as described by building your custom image? Installing apps directly into a running container isn’t persistent and can lead to issues like the one you’re experiencing.

[NeihTzxc]

The whole point of Docker is to make environments machine-independent, so this issue shouldn’t be caused by your PC.

Did you install the applications as described by building your custom image? Installing apps directly into a running container isn’t persistent and can lead to issues like the one you’re experiencing.

I've tried rebuilding it, but it's sometimes faulty and sometimes not. I haven't been able to investigate the cause yet.

[NeihTzxc]

The whole point of Docker is to make environments machine-independent, so this issue shouldn’t be caused by your PC.

Did you install the applications as described by building your custom image? Installing apps directly into a running container isn’t persistent and can lead to issues like the one you’re experiencing.

I've looked up the possible cause here:
#1555
I tried #1555 (comment) and it worked.
I still don't understand why this bug has been around for so long but hasn't been fixed yet.

[DanielRadlAMR]

Is this a setup that has been running for over a year? There might still be some leftover artifacts causing issues.

I’d recommend cleaning up your setup and recreating all containers. This bug no longer exists in the current version, but you may still be affected by previously faulty versions. In general, it’s a good idea to do at least a minimal cleanup each time you deploy a new image.

[jslocomotor]

I've had this happen on one of my test instances when updating from 3.0.0 to 3.0.1 as well. Tried everything to fix it, spent 30 minutes on it and then dumped the instance. When migrating multiple production instances from 2.x.x to the latest 3.2.0, however, everything went well.
If there is some potential for deployments to break when upgrading to 3.2.0 I would suggest migrating just the site using backup and restore to new, clean volumes.
Do we have documentation for this? Would be quite the same as server migration:
Backup site with files, export from volume, clear volumes / migrate to new server, start minimal stack to create volumes, import DB and site?

[DanielRadlAMR]

The v3 versions had some bugs prior to #1894, which may have caused the issues you experienced.

Since that PR was merged, those problems should be resolved—this also aligns with your experience upgrading your production setups.

That said, there could still be issues related to older configurations, such as stale volumes being mounted unexpectedly. It might be worth recreating the containers and carefully reviewing any persisted data or volumes to ensure nothing outdated is being reused.