Provide fast feedback when Docker Daemon is not running

[jlink]

When Docker daemon is not running it can take quite a while for a test failing with exception. Moreover, to the un-initiated it might not be obvious that they must start a docker daemon.

[gtrefs]

I get that and look into it. I think the issue is with Testcontainers itself. I am not sure if there is a faster way to detect if the daemon is up.

[gtrefs]

What I see from the testcontainers code, there are several strategies in order to detect if docker is available. The following strategies are tried once the docker client should be used:

org.testcontainers.dockerclient.EnvironmentAndSystemPropertyClientProviderStrategy
org.testcontainers.dockerclient.UnixSocketClientProviderStrategy
org.testcontainers.dockerclient.DockerMachineClientProviderStrategy
org.testcontainers.dockerclient.NpipeSocketClientProviderStrategy
org.testcontainers.dockerclient.RootlessDockerClientProviderStrategy

Each of these strategies is tried by trial and error. If all strategies fail, an exception is thrown. I think this is where testcontainers spends most of its time. Loading the docker client eagerly might give you some extra seconds, but I am not sure if that is what you want. Also, I could add an extra log which is a bit more specific. What do you think?

[jlink]

If checking cannot be sped up maybe the feedback could be more revealing. Something like:
"No docker environment available. Please check that you're running Docker Desktop or ...."