Load test: set threshold for latency metrics and add periodic lists

Revert "Added e2e test case for network partition."

Decrease threshold for latency metrics in density tests

Increased failCount param in RunRC function for e2e tests.

The previous value was 10 (equivalent of 50sec) current is 24 (2min).
Time in brackets mean: how long should I wait for the next pod of RC to be created.

While we are creating 3000 pods for scalability tests it tends to fail.

Creating RCs sequentially not in parallel in Load e2e test

Ginkgo naming conventions (fixed)

Revert "Make services e2e test more debuggable"

Make services e2e test more debuggable

Prevent access to the /run and /exec endpoints on the master's kubelet

Insert 'svc' into the DNS search paths

Reduce number of list requests in watch-based cache

Use Node IP Address instead of Node.Name in minion.ResourceLocation.

Kubelet: Add metric for tracking Docker error count

Add an explicit variable to indicate whether an instance is master or not

Kubelet unit test cleanup

Fix a bug in kubectl exec handling.

This reverts commit 479e5005, reversing
changes made to 5115f0e6.

Revert "Cleanup ginkgo declarative naming in E2Es"

Refactor GetNodeHostIP into pkg/util/node (instead of pkg/util to break import cycle).

Include internalIP in gce NodeAddresses. Remove NodeLegacyHostIP

Add a new contiguous allocator strategy option

Add an advertise-address flag

Updated node.md

Explain that file-based pods cannot use secrets

Don't assume that CA_CERT_BASE64 will be set.

Cleanup ginkgo declarative naming in E2Es

Allows the tracking of errors by Docker operation.