* Upgraded Kubernetes from 1.5.4 to 1.7.3
* Upgraded Calico from 0.23.0 to 2.4.1
* Upgraded kube-dns from 1.9 to 1.14.4
* Upgraded dns-autoscaler from 1.0.0 to 1.1.2
* Upgraded heapster from 1.2.0 to 1.4.1
* Upgraded kube-dashboard from 1.5.0 to 1.6.3
* Changed from using etcd2 to etcd version 3
* Changed Calico to the default network solution
* Changed POD_NETWORK from 10.2.0.0/16 to 192.168.0.0/16 as Calico does not respect CALICO_IPV4POOL_CIDR
* Changed Vagrant memory limits for the controller, increased to 2Gb
* Removed Kubelet usage of deprecated flag --api-servers
* Removed Kubelet usage of deprecated flag --register-schedulable
* Added a kubeconfig to make up for the removed flags above
* Added possibility to modify how many cpu cores Vagrant assigns the hosts

* Upgraded Kubernetes from 1.7.3 to 1.7.4
* Upgraded Calico from 2.4.1 to 2.5.0
* Reverted configuration change for POD_NETWORK as my understanding of
  how Calico treats CALICO_IPV4POOL_CIDR was incorrect.
  CALICO_IPV4POOL_CIDR is used to create an initial pool and cannot
  later be changed and used as a way to assign a new pool without first
  removing the existing pool(s). Thanks @tmjd for clarifying this!