mirror of
https://github.com/plankanban/planka.git
synced 2026-02-05 00:39:58 +03:00
Websocket connection error #506
Reference in New Issue
Block a user
Delete Branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @Oscar-Charlie on GitHub (May 13, 2024).
Hello together,
We have setup planka in our company via docker on our synology.
Configuration is as follows:
version: '3.9'
services:
planka:
image: ghcr.io/plankanban/planka:latest
container_name: Planka
command: >
bash -c
"for i in
seq 1 30; do./start.sh &&
s=
? && break || s=?;echo "Tried $$i times. Waiting 5 seconds...";
sleep 5;
done; (exit $$s)"
restart: on-failure:5
volumes:
- /volume1/docker/planka/avatars:/app/public/user-avatars
- /volume1/docker/planka/images:/app/public/project-background-images
- /volume1/docker/planka/files:/app/private/attachments
ports:
- 3617:1337
environment:
- BASE_URL=https://planka.ourwebsite.com
- TRUST_PROXY=1
- DATABASE_URL=postgresql://postgres@postgres/planka
- SECRET_KEY=xxxxx
- DEFAULT_ADMIN_EMAIL=xxx@email.de
- DEFAULT_ADMIN_PASSWORD=xxxx
- DEFAULT_ADMIN_NAME=xxxx
- DEFAULT_ADMIN_USERNAME=xxxx
- NODE_ENV=production
postgres:
image: postgres
container_name: Planka-DB
restart: on-failure:5
volumes:
- /volume1/docker/planka/db:/var/lib/postgresql/data:rw
environment:
- POSTGRES_DB=planka
- POSTGRES_HOST_AUTH_METHOD=trust
We have set a reverse proxy rule to 443 and to localhost 3617. We also set the websocket option in our synology.
Nevertheless one of our client just sees the loading screen from the startup an can't get any further. From the google debug consule they get an websocket error (sails.io)
Do you have any advice? We are loving planka, but we need to have to make it work for all of our clients.
Thank you in advance,
Matthias
@meltyshev commented on GitHub (May 14, 2024):
Hi! It seems that your settings look correct. Please make sure that
BASE_URLis exactly the same as the address you specify in your browser and also starts withhttp://orhttps://. It's hard for me to suggest solutions to possible server-side problems, as I'm not very familiar with it, but I'll try to find a way to display an error text in the server logs.@meltyshev commented on GitHub (May 14, 2024):
Please try to change
commandindocker-compose.ymlfromto
With this you'll be able to run Planka in development mode and see all debug messages.
Then try to open Planka in a browser and get that socket connection error. Check the logs on the server, there should be something like "debug: A socket was rejected via the ...". Also, there may not be any error there, but then the problem is somewhere in the proxy server settings.
@modem7 commented on GitHub (Nov 7, 2024):
I'm getting the same issue on this end.
It works if I set the container to load locally (bypassing the reverse proxy).
@lid3rs commented on GitHub (Jan 22, 2025):
Have the same error, using reverse proxy
@pentiumoverdrive commented on GitHub (Feb 5, 2025):
I have web sockets issue as well, but only through Microsoft Entra Application Proxy.
Websocktes works fine when accessing planka on the server that runs the Entra Application Proxy connector, but comming over internet through the reverse proxy there is just an infinite spin after login.
It apperas that Entra Application Proxy does not support web sockets fully, but other applications that uses web sockets works at least.
@ghost commented on GitHub (Feb 7, 2025):
Same here, and it's just about driven me mad troubleshooting and debugging.
@lid3rs commented on GitHub (Mar 17, 2025):
Any plans fixing this?
@ghost commented on GitHub (Apr 23, 2025):
I'm having the same issue, and have been unable to resolve it.
@00pp commented on GitHub (Jun 11, 2025):
same issue. can't resolve
@TheGitPanda commented on GitHub (Jun 17, 2025):
Synology Docker users, head to Control Panel > Login Portal > Reverse Proxy > [Edit the proxy config] > Custom Header > Create > Websocket. It should insert two rows;
Upgrade : $http_upgrade
Connection: $connection_upgrade
This solved it in my case.
For the record, I had tweaked some files on the planka docker container prior to this (which I'm intending to rollback when I have time)
$ vi /app/config/sockets.jsEnsuring beforeConnect was enabled and return
process(undefined, true)Docs: https://sailsjs.com/documentation/reference/configuration/sails-config-sockets
$ vi /app/config/security.jsEnsuring
allowOrigins: '*', allowAnyOriginWithCredentialsUnsafe : trueDocs: https://sailsjs.com/documentation/reference/configuration/sails-config-security
Hope this helps
@S1M8N commented on GitHub (Jul 19, 2025):
@TheGitPanda thank for the solution it's work fine with cloudflared tunnel environnement
@pvdh commented on GitHub (Jan 11, 2026):
I had the very same issue with planka and traefik.
Turns out: I had seafile running on the same server and following a seafile-traefik-guide added
"traefik.http.routers.seafhttp.rule=Host(
seafile.$DOMAINNAME) && PathPrefix(/sdoc-server) || PathPrefix(/socket.io)"for seafile. Which is evaluated as (Host(
seafile.$DOMAINNAME) && PathPrefix(/sdoc-server)) || PathPrefix(/socket.io) and therefore conflicts with /socket.io in planka.Changing the Rule to "Host(
seafile.$DOMAINNAME) && (PathPrefix(/sdoc-server) || PathPrefix(/socket.io))" solved my issue.