Websocket error, services not starting


I’ve been running cells for quite a while now and it’s been great. A few days ago I started getting this error on the UI and the Workspaces aren’t syncronizing:
Websocket seems to be disconnected: please make sure that the server is up and websocket connection is working
I’m running Cells 3.0.9 on a v-server from Strato (no docker) as a systemd service with Debian 10.
I’ve done some research and it seems to be related to a reverse proxy, I’m running it behind nginx version 1.20.0 . I had upgraded to the new version and thought that maybe this was the problem so I downgraded back to 1.20.0 .
I copied the config from the instructions and this is my nginx cells config file right now:

server {
    server_name cells.mydomain.com;

    # Allow any size file to be uploaded.
    client_max_body_size 0;
    # To disable buffering
    proxy_buffering off;

    location / {
        # Uncomment this to enable gRPC and thus be able to use cells-sync
        #if ($http_content_type = "application/grpc") {
        #    grpc_pass grpcs://cells:8080;
        proxy_pass https://localhost:8080;

    location /ws/ {
        proxy_pass https://localhost:8080;
        proxy_http_version 1.1;
        proxy_set_header Upgrade $http_upgrade;
        proxy_set_header Connection "Upgrade";

    error_log /var/log/nginx/cells-proxy-error.log;
    access_log /var/log/nginx/cells-proxy-access.log;

    listen [::]:443 ssl;
    listen 443 ssl http2;

    # certificate configuration (in this case generated by certbot)
    ssl_certificate /etc/letsencrypt/live/cells.mydomain.com/fullchain.pem;
    ssl_certificate_key /etc/letsencrypt/live/cells.mydomain.com/privkey.pem;
    include /etc/letsencrypt/options-ssl-nginx.conf;
    ssl_dhparam /etc/letsencrypt/ssl-dhparams.pem;


server {
        if ($host = cells.mydomain.com) {
            return 301 https://$host$request_uri;

        listen 80;
        listen [::]:80;
        server_name cells.mydomain.com;
        return 404;

When I go to the data sources I see “Node undefined not found”

This is what the data source looks like (I’ve checked, the owner of the folder is the pydio user):

The nginx proxy error log doesn’t give any useful info

cells ps just prints out a warning that no private IP was detected for binding broker but doesn’t show the services, most of which aren’t running as I can see from the UI.

Does anyone have any ideas?



I have absolutely no idea if this helps, but at some point in the past I really had some difficulties getting Pydio Cells to ‘attach’ internally to localhost, as you’re doing (aye, I use nginx as a reverse proxy as well).

Somewhere (where?) I found a warning that localhost, for some obscure reason, would not work with Pydio Cells; instead, one should assign a new, separate, private IP address.

You might have a look at your /etc/network/interfaces (or possibly any file inside /etc/systemd/network/) and add an extra private IP address to your default network device (I’m assuming Ethernet, or some sort of virtual Ethernet), and configure Pydio Cells to use that one instead — and then change the nginx configuration, too.

I’m afraid that I can’t post a working configuration for Debian 10; currently, my production Cells is running under Ubuntu, and the guys from Canonical pushed all the network configurations to use Netplan instead…

Hey, thanks for the quick response. Unfortunately my provider says that the network is controlled through Hypervisor so it’s not possible to add another LAN IP :frowning: Pretty annoying

by the way, this is what the firefox console shows

Firefox can’t establish a connection to the server at wss://cells.mydomain.info/ws/event. 2 [login line 1 > injectedScript line 1 > eval:1:3604473](https://cells.mydomain.info/login line 1 > injectedScript line 1 > eval)

WebSocket Closed Connection:The connection was closed abnormally, e.g., without sending or receiving a Close control frame (code 1006) [login line 1 > injectedScript line 1 > eval:1:3931532](https://cells.mydomain.info/login line 1 > injectedScript line 1 > eval)

Error: Internal Server Error

Request https://cells.mydomain.info/login line 1 > injectedScript line 1 > eval:1

emit https://cells.mydomain.info/login line 1 > injectedScript line 1 > eval:1

onreadystatechange https://cells.mydomain.info/login line 1 > injectedScript line 1 > eval:1

[login line 1 > injectedScript line 1 > eval:1:42907](https://cells.mydomain.info/login line 1 > injectedScript line 1 > eval)

Basically the same error repeats for wss://cells.mydomain/ws/chat

I’d love to hear ideas if anyone has any.


Just to get Nginx out of the equation I switched to Apache reverse proxy, I get the exact same errors :frowning:

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.