Added pending phase for volumes.  added defaults for PV/PVC.  refactored to better phase transitioning in control loops

Fixed typo in getting started documentation

Pass the CA root cert into the apiserver

Make argument order for testContainerOutputInNamespace consistent

It fixes a regression in fedora guide and provides minor updates.

Fetch the kibana tar.gz over SSL.

Fixes in e2e tests: checking pods in the right namespace; proper use of namespaces in docker container test.

Fix crash in kubelet when persistent volume claim is not bound.

Fixed URLs in docker getting started guide.

Corrected some typos in docs

While testing #6105 ran into this issue, kubelet crashed because
controller had not yet bound the claim to a volume.

Is there a docker run command to build this?

kubelet/metrics: Move instrumented_docker.go to dockertools.

This can solve the circular import problem when we move the
kubelet.pullImage to kubelet/metrics or kubelet/container package.

perform client cert checks for authorization. Only enable on GCE where
the apiserver is terminating SSL connections from end users.

Address a couple of TODO's in the recent rolling update change.

Specify host/container ports for InfluxDB, so that the UI is accessible

Warning about Critical bug in the GlusterFS Volume Plugin

This warning will serve to let users know about a critical bug so they don't experience it first hand - https://github.com/GoogleCloudPlatform/kubernetes/issues/7317

Kubelet: filter out terminated pods in SyncPods

First part of improved rolling update, allow dynamic next replication controller generation.

Once a pod reaches a terminated state (whether failed or succeeded), it should
not transit out ever again. Currently, kubelet relies on examining the dead
containers to verify that the container has already been run. This is fine
in most cases, but if the dead containers were garbage collected, kubelet may
falsely concluded that the pod has never been run. It would then try to restart
all the containers.

This change eliminates most of such possibilities by pre-filtering out the pods
in the final states before sending updates to per-pod workers.

Raise cAdvisor stats cache to 2m.