Cluster admin repeated websocket error 1006

[DavidePrincipi]

Since 2024-04-11 new installations of NS8 have a repeated cluster-admin websocket disconnection with error code 1006.

Steps to reproduce

• Install NS8 core 2.6.0

Expected behavior

I can connect to cluster-admin web UI

Actual behavior

The UI repeatedly is disconnected.

In the browser console

Websocket disconnected 
close { target: WebSocket, isTrusted: true, wasClean: false, code: 1006, reason: "", srcElement: WebSocket, currentTarget: WebSocket, eventPhase: 2, bubbles: false, cancelable: false, … }

...and consequent failures.

Components

• core 2.6.0 or testing
• any browser
• any client OS

[DavidePrincipi]

It seems the issues comes from a recent Traefik release.

[traefik1@rl1 state]$ podman images                                       
REPOSITORY                 TAG         IMAGE ID      CREATED       SIZE   
docker.io/library/traefik  v2.11.1     cc12b4a7ccda  12 hours ago  161 MB 
docker.io/library/traefik  v2.11       cc12b4a7ccda  12 hours ago  161 MB 
docker.io/library/traefik  v2.11.0     c1ef9171e9d1  3 weeks ago   161 MB 

The issue is not reproducible with Traefik 2.11.0

Workaround

 bash install.sh ghcr.io/nethserver/traefik:latest

[DavidePrincipi]

Fixed in NethServer/ns8-traefik#ccb4ad42468b0ffde0d5d6f1eace7944c7212f8c

[DavidePrincipi]

Released in https://github.com/NethServer/ns8-traefik/releases/tag/2.2.1