v2.11.1 breaks websockets for asp.net core

[clarity99]

Welcome!

• Yes, I've searched similar issues on GitHub and didn't find any.
• Yes, I've searched similar issues on the Traefik community forum and didn't find any.

What did you do?

After I upgraded to the latest on 2.11 (.1) websockets keep closing automatically. I have an asp.net core mvc app with Blazor components which use websocket connections. In console I saw: Error: Connection disconnected with error 'Error: WebSocket closed with status code: 1006 (no reason given).'. After reconnect, it immediately closed again.

What did you see instead?

Error: Connection disconnected with error 'Error: WebSocket closed with status code: 1006 (no reason given).'.

What version of Traefik are you using?

v2.11.1

What is your environment & configuration?

command:
      - "--log.level=INFO"
      - "--api.dashboard=true"
      - "--entrypoints.web.address=:80"
      - "--entrypoints.websecure.address=:443"
      - "--providers.docker=true"
      - "--providers.docker.exposedbydefault=false"
      - "--certificatesresolvers.letsencrypt.acme.httpchallenge=true"
      - "--certificatesresolvers.letsencrypt.acme.httpchallenge.entrypoint=web"
      - "--certificatesresolvers.letsencrypt.acme.email=xxx"
      - "--certificatesresolvers.letsencrypt.acme.storage=/letsencrypt/acme.json"

Add more configuration information here.

If applicable, please paste the log output in DEBUG level

No response

[clarity99]

downgrading to v2.11.0 solved the issue.

[ldez]

I think it's related to #10569.

The related documentation: https://doc.traefik.io/traefik/migration/v2/#v2111

can you try to set lingeringTimeout to 0?

[Majestic7979]

I want to add that I found this after suspecting issues with some applications I host that use websockets. Vaultwarden was having issues - the desktop app would not stop refreshing and syncing, and the notifications that require websockets were not working. Home Assistant would also constantly refresh, it was driving me mad. Adding this "lingeringTimeout" section and settings to 0 on my web and web-secure entryPoints fixed the issue:

entryPoints:
  web:
    address: ":80"
    transport:
      respondingTimeouts:
        tcp:
          lingeringTimeout: 0

  web-secure:
    address: ":443"
    transport:
      respondingTimeouts:
        tcp:
          lingeringTimeout: 0

[RowanTaubitz]

Same here with homeassistant, constant refreshing of the dashboard and phone/tablet applications throughout the house!
respondingTimeouts.TCP.lingeringTimeout=0 worked to fix the issue

[milindpatel63]

i'm also having this issue.

[comassky]

Same issue, websockets are broke

Works with v2.10, broken since 2.11

Works in 2.11 & V3.0-rc4 with lingeringTimeout: 0

[dabeck]

Why would one introduce such a breaking change in a patch version 😩

[echoblag]

Adding to the list of reported issues, updating to 2.11.1 broke HTTP connections on one of my WordPress sites. Constant connection interrupts, couldn't find any traces or debug triggers, finally downgraded back to Traefik 2.11.0 and that fixed it.

Looks like this is more than just websockets.

I will be spending some time later looking at the lingering timeout config.

[mbleichner]

This also broke our instance of Mattermost - seeing a lot of connection errors in chrome devtools. Reverting to 2.11.0 fixed it.

[thespad]

It's one thing to add this in the v3 RCs, it's quite another to do it as part of a patch release to the stable branch.

The note of

The new default value is purposely narrowed and can close the connection too early.

Is even worse, because it means you knew it was likely to be a breaking change in a lot of cases.

[Twiederh]

I added LingeringTimeout to my traefik.yml:

` entryPoints:
http:
address: ":80"
transport:
respondingTimeouts:
tcp:
lingeringTimeout: 0
http:
redirections: # HTTPS redirection (80 to 443)
entryPoint:
to: "https" # The target element
scheme: "https" # The redirection target scheme

https:
address: ":443"
transport:
respondingTimeouts:
tcp:
lingeringTimeout: 0`

My NodeRed dashboard is working again, however my Homeassistant instance still has the same issue that the page is reloading every 5 seconds or so.
I went back to v2.11.0 for the time being until this is completely resolved.

[ghhv]

Same for ASP.NET 8 Blazor Server (& Dozzle by the way) - switching back to 2.11.0 resolved. A day wasted to get to this point 😢😫🤬🫨

[emilevauge]

On behalf of the maintainer team, sorry for the inconvenience.
We opened a global issue to cover this bug: #10598
I'm closing this one, as we will update the global issue while we make progress on this.

[davidnewhall]

Hi, I'm just a nobody that's never used this software. Friendly advice: If your goal is to get Fortune 500 companies to run your software, avoid referring to their lost revenue as an inconvenience.