Issues with Testcontainers' ryuk not shutting down since update to v.17.0/v1.17.1

[daviian]

Actual Behavior

Testcontainers' data reaper container Ryuk should shut itself down when the connection from the test process is closed.
When using IPv4 though, Ryuk does not notice that the connection got closed. The problem does not occur when IPv6 is used.

Steps to Reproduce

Following scenario outlines the issue:

# Start Ryuk
docker run -v /var/run/docker.sock:/var/run/docker.sock -p 8080:8080 testcontainers/ryuk:latest

# Connect to Ryuk
telnet 127.0.0.1 8080

Close telnet session so that Ryuk would shut down.

Result

When closing the telnet session, Ryuk does not shut itself down as expected and is running forever.

Expected Behavior

Ryuk shuts itself down when noticing the closed connection.

Additional Information

As discussed and verified in the issue testcontainers/testcontainers-dotnet#1360 the root cause stems from most probably a regression bug that was introduced in v1.17.0. Rancher Desktop in version v.16.0 does not have this issue.

Rancher Desktop Version

1.17.1

Rancher Desktop K8s Version

Kubernetes is disabled

Which container engine are you using?

moby (docker cli)

What operating system are you using?

Windows

Operating System / Build Version

Windows 11

What CPU architecture are you using?

x64

Linux only: what package format did you use to install Rancher Desktop?

None

Windows User Only

No response

[jandubois]

Thank you for the very simple repro commands!

I've confirmed (on Windows 11) that with Rancher Desktop 1.17.1 the container never receives the EOF indicating that the connection has closed. This is indeed a regression from 1.16.0.

With the latest CI build I can confirm that the issue has been resolved (probably by upgrading gvisor-tap-vsock to 0.8.3):

I'm going to close this issue; the fix will be included in the 1.18 release.