Automatic merge from submit-queue

Setup fluentd-ds-ready label in startup script not in kubelet

fix #42840

``` release-note
TBD
```

Automatic merge from submit-queue

Relax 'misc' container memory constraints

Fixes https://github.com/kubernetes/kubernetes/issues/40607

/cc @dchen1107

Automatic merge from submit-queue

dns_autoscaling test: Don't expect configMap to be re-created with previous params

Fix #42766.

In dns_autoscaling e2e test, there is a test case that we delete the kube-dns-autoscaler configMap first and wait for it to be recreated with previous params. This is not necessarily true, because we will preserve the kube-dns-autoscaler configMap during upgrade but we may update the default params in kube-dns-autoscale binary, which means a configMap with new params will be created in the upgrade case.

One of the test failures could be found on: https://k8s-gubernator.appspot.com/build/kubernetes-jenkins/logs/ci-kubernetes-e2e-gke-gci-1.5-gci-1.6-upgrade-cluster-new/18#k8sio-dns-horizontal-autoscaling-kube-dns-autoscaler-should-scale-kube-dns-pods-in-both-nonfaulty-and-faulty-scenarios

This PR removes the codes that make above assertion. 

@bowei @skriss

```release-note
NONE
```

Automatic merge from submit-queue

kubectl: fall back to computing QoS class if server does not populate

Fixes https://github.com/kubernetes/kubernetes/issues/42831

https://github.com/kubernetes/kubernetes/pull/38989 changed how kubectl determines the the QoS Class field.  In order to prevent reimplementation of the QoS class logic in each client, the API server was modified to set the QoS Class in a new status field.  However API servers 1.5 and older do not set this field.  This results in a kubectl 1.6 client talking to a 1.5 API server to not print the QoS class field, which is a regression.

This PR causes kubectl to fall back to the old method of evaluting the QoS class when the status field from the API server is not set.

@vishh @derekwaynecarr @ymqytw @pwittrock @skriss

Automatic merge from submit-queue

Allow ABAC to be disabled easily on upgrades

**What this PR does / why we need it**:
Adds a local variable to the configure-helper script so that ABAC_AUTHZ_FILE can be set to a nonexistent file in kube-env to disable ABAC on a cluster that previously was using ABAC.

@liggitt @Q-Lee

Automatic merge from submit-queue

Update npd to the official v0.3.0 release.

Update npd to the official release v0.3.0.

This also fixes a npd bug https://github.com/kubernetes/node-problem-detector/pull/98.

@dchen1107 @kubernetes/node-problem-detector-reviewers

Automatic merge from submit-queue

Add guards for StatefulSet and AppArmor upgrade testing

This PR adds automated upgrade infrastructure to allow test suites to know what versions and node images are going to be testing and whether or not they should be skipped. It also adds a guard to prevent StatefulSets from being tested with versions prior to 1.5.0, and a guard to prevent AppArmor from running on distros other than gci and ubuntu.

Automatic merge from submit-queue

Retry kubectl test replace on conflict

Since kubectl is doing a resource-version-constrained replace, it is subject to conflicts on a contentious resource (like a node managed by the node controller)

Fixes #41892 (the specific flake, not the watch cache issue)

Automatic merge from submit-queue (batch tested with PRs 43177, 43202)

Rename default storageclasses

From UX perspective, 'default' is a bad name for the default storage class:

```
$ kubectl get storageclass
NAME                TYPE
default (default)   kubernetes.io/aws-ebs
```

This is sort of OK, it gets more confusing when user is not happy with the
preinstalled default storage class and creates its own and makes it default:

```
NAME             TYPE
default          kubernetes.io/aws-ebs
iops (default)   kubernetes.io/aws-ebs
```

This PR uses name of the underlying storage as name of the default storage class:

```
NAME            TYPE
gp2 (default)   kubernetes.io/aws-ebs
```

On GCE (and many others):
```
NAME                 TYPE
standard (default)   kubernetes.io/gce-pd
```

Detailed list of names of new default storage classes:

* AWS: `gp2`
* GCE: `standard` (from pd-standard)
* vSphere: `thin`

* Cinder does not have a default - it's up to OpenStack admin to set some default and it can change at any time, using `standard` as the class name.
* I was not able to find details about Azure, using `standard` too.

@justinsb @jingxu97 @kerneltime @colemickens, PTAL quickly so we can catch 1.6.

```release-note
NONE
```

For 1.6 release manager, this PR just renames objects in addon manager.

Automatic merge from submit-queue

Force create symlinks.

Followup from #43070.  Fixes #42660 in a way that does not break etcd upgrades.

Automatic merge from submit-queue

Bump cluster autoscaler to 0.5.0-beta2

**What this PR does / why we need it**:

This part is a part of Cluster Autoscaler release process for 1.6.0. It contains couple of bugfixes on top of 0.5.0-beta1. Hopefully this will be the last beta before final bump to 0.5.0

cc: @MaciekPytel @jszczepkowski @fgrzadkowski

Automatic merge from submit-queue

recycle pod can't get the event since channel closed

What this PR does / why we need it:
We create a   hostPath type  PV with "Recycle" persistentVolumeReclaimPolicy,  and bind a PVC to it, but after deleted the PVC, the PV cannot become to available status. This is happened after we upgrade etcd to 3.0. The reason is:
If the channel used to get the pod message and events been abnormal closed(for example, the event channel maybe closed because of "required revision has been compacted" error), the function internalRecycleVolumeByWatchingPodUntilCompletion will stuck in a loop, and the recycle pod will not been deleted, the PV can not become into available status

Special notes for your reviewer:
None
Release note:

From UX perspective, 'default' is a bad name for the default storage class:

$ kubectl get storageclass
NAME            TYPE
default (default)   kubernetes.io/aws-ebs

This is sort of OK, it gets more confusing when user is not happy with the
preinstalled default storage class and creates its own and makes it default:

NAME             TYPE
default          kubernetes.io/aws-ebs
iops (default)   kubernetes.io/aws-ebs

Automatic merge from submit-queue (batch tested with PRs 43122, 43086)

Configure example pod namespaces

Fixes #43015 
Fixes #43019
Fixes #38222

Automatic merge from submit-queue

Prevent protobuf storage with etcd2

Prevents accidentally storing protobuf content in etcd2 when upgrading to 1.6

c.f. https://github.com/kubernetes/kubernetes/issues/42976#issuecomment-286537139

```release-note
if kube-apiserver is started with `--storage-backend=etcd2`, the media type `application/json` is used.
```

Automatic merge from submit-queue (batch tested with PRs 43180, 42928)

Fix waitForScheduler in scheduer predicates e2e tests

**What this PR does / why we need it**: Fixes waitForScheduler in e2e to resolve flaky tests in scheduler_predicates.go

**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #42691

**Special notes for your reviewer**:

**Release note**:

```release-note
NONE
```

Automatic merge from submit-queue (batch tested with PRs 43180, 42928)

Verify generated protobuf script should fail on staging/ changes too

fixes #35486

Automatic merge from submit-queue (batch tested with PRs 43162, 43157)

Use beta default class annotation for default storageclass tests.

**What this PR does / why we need it**:
The default storageclasses are still installed with the beta annotation, so the test should explicitly use the beta annotation.

**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #43150

**Special notes for your reviewer**:

**Release note**:

```release-note
NONE
```

Automatic merge from submit-queue

Add saad-ali and marun to test/OWNERS

/assign @saad-ali @marun 

Also ensure that approvers are in the reviewer list, and sort both lists.

And convert StatefulSet's version skipping to the new API.

Automatic merge from submit-queue (batch tested with PRs 40964, 42967, 43091, 43115)

Add process debug information to summary test

Print out the processes in each system cgroup when the Summary API test fails, to help debug https://github.com/kubernetes/kubernetes/issues/40607

/cc @yujuhong @Random-Liu

Automatic merge from submit-queue (batch tested with PRs 40964, 42967, 43091, 43115)

fixes dswp flake

Sometimes a pod may not appear in desired state
of world immediately, we poll before failing.

It only adds additional 30s to tests in worst case.

Fixes #42990 

cc @jingxu97

Automatic merge from submit-queue (batch tested with PRs 40964, 42967, 43091, 43115)

Update hack scripts to use godep v79 and ensure_godep_version

**What this PR does / why we need it**:
Based on #42965 and https://github.com/kubernetes/kubernetes/pull/42958#discussion_r105568318, this pins the godep version at v79, which should fix some issues when running godep in go1.8 local environments.

**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #42817

**Special notes for your reviewer**:
This should likely get the v1.6 milestone so that it can be merged into master. While I'm setting a default godep version, I'm continuing to use the local pins per this comment: https://github.com/kubernetes/kubernetes/pull/42965#issuecomment-285962723 .

**Release note**:

```release-note
NONE
```

cc: @sttts

Automatic merge from submit-queue (batch tested with PRs 40964, 42967, 43091, 43115)

Improve code coverage for pkg/kubelet/status/generate.go

**What this PR does / why we need it**:

Improve code coverage for pkg/kubelet/status/generate.go  from #39559

Thanks.

**Special notes for your reviewer**:

**Release note**:

```release-note
```

Automatic merge from submit-queue

Guarantee watch before action in e2e event observer helper function.

**What this PR does / why we need it**:

Adds a missing synchronization barrier to an e2e event observation helper function.

- This change should guarantee that in observeEventAfterAction,
  the action is only executed after the informer begins watching
  the event stream.

**Release note**:

```release-note
NONE
```

cc @kubernetes/sig-scheduling-pr-reviews @bsalamat