due to this commit:
991b5f6

need a workaround if we don't use docker-machine to start it.

Create a travis or circleci CI, and automatically push new release when k8s gets released...

• kubectl commands often time out. E.g., kubectl create -f frontend-controller.yaml
  error: could not read an encoded object from frontend-controller.yaml: unable to connect to a server to handle "replicationcontrollers": couldn't read version from server: Get http://localhost:8080/api: EOF
• Works more consistently if I give a second 'host-only' network to the boot2k8s VM (configure VBox DHCP to give it 192.168.64.100) and use that IP as the target for kubectl
• Downloading of images is flaky. I wonder if giving the boot2k8s a proper disk (like boot2docker /dev/sda1) will help.
• As of now, it is quite unusable

From kubelet.sh file set cluster DNS to 10.0.0.10
--cluster-dns=10.0.0.10

So, I would like to run SkyDNS as a service discovery, but the service cluster ip range set to 10.200.20.0/24
--service-cluster-ip-range=10.200.20.0/24

The error message as below:
The Service "kube-dns" is invalid.
spec.clusterIP: Invalid value: "10.0.0.10": provided IP is not in the valid range

FYI: I'm running boot2k8s on VMware Fusion on OS X.

While we use this for kmachine.
We also need to make sure it works standalone, and that uploaded releases work as is.

If you run kubectl port-forward and try to access the port it will fail with the error
"unable to do port forwarding: socat not found." and indeed it is not installed on the iso.

Unfortunately I was not able to use tce_load to install socat because the 64-bit version of socat is missing from the tinycore repos for some reason. (You can access http://tinycorelinux.net/6.x/x86/tcz/socat.tcz but not http://tinycorelinux.net/6.x/x86_64/socat.tcz)