Bug: NS_ERROR_CORRUPTED_CONTENT when visiting zipline proxied via nginx, which is behind cloudflare?

[pngmeow]

Bug description

I’m currently experiencing issues with the Zipline dashboard. The dashboard takes a long time to load, and in some cases, it does not load at all.

While inspecting the issue in Firefox DevTools, I noticed that several JavaScript modules are failing with the NS_ERROR_CORRUPTED_CONTENT error.

However, when I check the affected modules directly on the server, the files exist and their contents appear to be valid and intact.

How is Zipline being run?

On docker (docker, docker compose, etc.)

Zipline Version

trunk (latest)

If applicable, what browsers are you seeing this issue on?

Firefox based (Firefox, Zen Browser, Waterfox, etc.)

Relevant Logs

No response

Steps to Reproduce

Steps to reproduce:

1. Install the latest trunk version of Zipline using Docker Compose.
2. Install NGINX and configure it as a reverse proxy for the Zipline dashboard.
3. Put the domain behind Cloudflare using the default Cloudflare settings.
4. Open the dashboard in Firefox.
5. Observe intermittent slow loading, failed page loads, or the NS_ERROR_CORRUPTED_CONTENT error.

[pngmeow]

Also worth to notice:

Even if the dashboard loads in, and we attempt to go to other part of the dashboard, we get hit with various errors, like these:

• Bug: hydration error #860
• Bug: Dashboard Client Error #866

I'm on the latest version of firefox, without any addons.

[diced]

could you provide your nginx conf? could you also try accessing zipline without using nginx, like through your ip directly. and doing the same without cloudflare etc. you can also try accessing on a chromium browser and seeing if it works or not. it's most likely some misconfigured nginx thing.

[pngmeow]

nginx.conf

#user http;
worker_processes auto;

error_log  logs/error.log;
error_log  logs/error.log  notice;
error_log  logs/error.log  info;

#pid        logs/nginx.pid;


# Load all installed modules
include modules.d/*.conf;

events {
    worker_connections  1024;
}

http {
        include /etc/nginx/mime.types;
        default_type application/octet-stream;

        client_body_timeout 300s;
        client_header_timeout 300s;
        send_timeout 300s;

        client_max_body_size 200m;
        client_body_buffer_size 1m;

        proxy_buffering on;
        proxy_buffer_size 64k;
        proxy_buffers 16 64k;
        proxy_busy_buffers_size 128k;

        sendfile on;
        keepalive_timeout 65;

        types_hash_max_size 4096;
        types_hash_bucket_size 128;

        gzip on;

        ssl_session_cache shared:SSL:50m;
        ssl_session_timeout 1d;
        ssl_session_tickets on;

        include /etc/nginx/sites-enabled/*;
}

sites-available/zipline.conf

server {
    listen 80;
    server_name u.foxtsukihime.top;

    return 301 https://$host$request_uri;
}

server {
    listen [::]:443 ssl;
    listen 443 ssl;
    server_name u.foxtsukihime.top;

    ssl_certificate /etc/letsencrypt/live/foxtsukihime.top/fullchain.pem;
    ssl_certificate_key /etc/letsencrypt/live/foxtsukihime.top/privkey.pem;
    include /etc/letsencrypt/options-ssl-nginx.conf;
    ssl_dhparam /etc/letsencrypt/ssl-dhparams.pem;

    client_max_body_size 100M;

    location / {
        proxy_pass http://127.0.0.1:3000;
        proxy_set_header Upgrade $http_upgrade;
        proxy_set_header Connection $http_connection;
        proxy_http_version 1.1;
        proxy_set_header Host $host;
        proxy_set_header X-Real-IP $remote_addr;
        proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
        proxy_set_header X-Forwarded-Proto $scheme;
    }
}

[pngmeow]

I dropped cloudflare, it seems to not do this anymore. Might be a CF issue, but I'm using default settings.
Beforehand I tested this with ungoogled chromium and it seemed to also throw this error.

[diced]

gzip on;

it's probably this, try turning gzip off

[pngmeow]

gzip on;

it's probably this, try turning gzip off

Turned it off, but the website still does the NS_CORRUPTED stuff (Yes, i restarted nginx)

[diced]

Turned it off, but the website still does the NS_CORRUPTED stuff (Yes, i restarted nginx)

have you tried enabling http2 in the zipline conf? try adding that in and seeing

[pngmeow]

Turned it off, but the website still does the NS_CORRUPTED stuff (Yes, i restarted nginx)

have you tried enabling http2 in the zipline conf? try adding that in and seeing

It is disabled.

[pngmeow]

Also, worth to note that the uploader itself works if i try to connect via a local IP, and there's no weird issues.

[diced]

Turned it off, but the website still does the NS_CORRUPTED stuff (Yes, i restarted nginx)

have you tried enabling http2 in the zipline conf? try adding that in and seeing

It is disabled.

ok yea, could you try enabling http2?