Docker rm returning ContainerNotFoundException after successfully removing container

Question

I am using a C# API that talks to DockerEngine (Docker.DotNet) to do some things, but notably, in this case, removed a stopped container. Here is my code:

foreach (var container in RunningContainers)
{
    if (container.State == "dead" || container.State == "exited")
    {
        long exitType = Client.Containers.InspectContainerAsync(container.ID).Result.State.ExitCode;

        if (exitType != 0)
        {
            badExitCodeCount++;
        }
        Client.Containers.RemoveContainerAsync(container.ID, new ContainerRemoveParameters()).Wait();
    }
}

Note that in this case RunningContainers still includes the stopped containers and hasn't had its state refreshed yet.

Here's the bug. The API successfully calls out to DockerEngine after executing RemoveContainerAsync, the container is removed from my local Docker Containers. Then shortly after the code moves on, a Docker "No such container" response is received, throwing an exception.

Any ideas about what's going on here?

Why do you use DockerEngine's async API in synchronous fashion? — Peter Csala, Jul 22 '21 at 11:29
@zkatancik I do believe that you can still use `async void` to be able to utilize `await` ([Reference](https://stackoverflow.com/questions/39260486/is-it-okay-to-attach-async-event-handler-to-system-timers-timer)) — Peter Csala, Jul 23 '21 at 06:12

score 0 · Answer 1 · answered Jul 23 '21 at 01:39

0

The problem lied within multiple threads running the method at the same time, tricky race condition I missed.

answered Jul 23 '21 at 01:39

zkatancik

1
3

Docker rm returning ContainerNotFoundException after successfully removing container

1 Answers1