Reclassify Restart tests as [Disruptive] and skip [Feature:.+] tests in -serial jobs

Log more information for ssh command.

Determinisitically get kubectl proxy port

E2E libraryzation (minimal patch)

Add cloudprovider methods for ssl.

Updating cAdvisor deps to v0.20.4

Update API docs.

Auto commit by PR queue bot

Restrict build jobs to running on build node instead of master

Add debug output to kubectl-log-e2e.

e2e reboot: prints stats and logs for not ready pods before and after the test

Refactor TestE2E to work correctly in parallel mode

Sometimes e2e test nodes are refused to reboot, but we didn't properly
catch the exit code. This one can give us more information on debugging.

Reclassify [Example] and [Feature:Example]

Fix a small error comment in nodecondition type

Fixed spelling typo

Augment logs in runContainer path

Fixes #20287.

Increase initial delay for liveness probe test pod from 15s to 30s.

Fix list-resources.sh to work with the GKE provider

make wget in gen-swagger-doc more resilient

Poll nodeip:node-port till kube-proxy updates iptables.

Fix e2e test failure for RecreateDeployment

The TestE2E function is run on every Ginkgo parallel node.
The cluster cleanup, sanity checking, and log dump operations are only
intended to be called once, however, so these have been moved out into
SynchronizedBeforeSuite and SynchronizedAfterSuite blocks.

Help debug port forward flakes

Auto commit by PR queue bot

Add more logging to diagnose flake

Remove provider skip lists

Bump glbc to 0.5.2

Auto commit by PR queue bot

Update Heapster API in Godeps