Automatic merge from submit-queue

Raise markVolMountedErr instead of unmountErr

Hit this when debugging https://github.com/kubernetes/kubernetes/issues/37657

We wrongly raised `unmountErr` when `markVolMountedErr` occurs.

Automatic merge from submit-queue

Begin paths for internationalization in kubectl

This is just the first step, purposely simple so we can get the interface correct.

@kubernetes/sig-cli @deads2k

Automatic merge from submit-queue (batch tested with PRs 38460, 39181)

Increase number of iterations in TestDefaulting

Automatic merge from submit-queue

remove wrapUpdatedObjectInfo from rbac storage

update by TODO

Automatic merge from submit-queue

kubectl: commas in --from-literal on secret creation

Closes #35185

``` release-note
Fixes an issue where commas were not accepted in --from-literal flags when creating secrets. Passing multiple values separated by a comma in a single --from-literal flag is no longer supported. Please use multiple --from-literal flags to provide multiple values.
```

Automatic merge from submit-queue (batch tested with PRs 39029, 39014)

[Glusterfs Vol Plugin]: Check kube client is invalid and return error

Fixes: #38939

In volume plugins, we need to create a kube client to make api call. And this kube client can be nil when, for example, wrong api-server configuration, but kubelet should not crash in this case.

I have also checked other plugins and found only glusterfs need this fix.

Automatic merge from submit-queue

Fix cloud-config name in test case

**What this PR does / why we need it**: fixes default cloud-config name in test cases for reset.

**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: related to kubernetes/kubeadm#75

**Special notes for your reviewer**:

**Release note**:
```release-note
NONE
```

Merge pull request #37887 from bruceauyeung/k8s-branch-use-bytes-buffer-instead-of-append-for-error-string-concat

Automatic merge from submit-queue

use bytes.Buffer instead of append for error string concat

**What this PR does / why we need it**:
1. in my benchmark test, `bytes.Buffer` takes much less time ( about 1:1000 ) than string append( `+=` ). 
>BenchmarkAppendConcat-4           100000            151438 ns/op          578181 B/op          2 allocs/op
BenchmarkBufferSprintf-4         3000000               487 ns/op              65 B/op          3 allocs/op
BenchmarkBufferConcat-4          5000000               271 ns/op              47 B/op          1 allocs/op

the benchmark codes is here  https://play.golang.org/p/LS52zGuwZN

2. in our `RunInitMasterChecks`, `RunJoinNodeChecks` there are lots of preflight checks. they may result in a huge error message. so `bytes.Buffer` can bring considerable performance enhancement in the worst of conditions.

beyond that, this PR 
1. fix an exported struct comment,
1. and use `found = append( found, errs...)` instead of for loop for simplicity.
Signed-off-by: bruceauyeung <ouyang.qinhua@zte.com.cn>

Automatic merge from submit-queue (batch tested with PRs 39093, 34273)

start breaking up controller manager into two pieces

This PR addresses: https://github.com/kubernetes/features/issues/88

This commit starts breaking the controller manager into two pieces, namely,
1. cloudprovider dependent piece
2. coudprovider agnostic piece

the controller manager has the following control loops -
- nodeController
- volumeController
- routeController
- serviceController
- replicationController
- endpointController
- resourceQuotaController
- namespaceController
- deploymentController 
  etc..

among the above controller loops,
- nodeController
- volumeController
- routeController
- serviceController

are cloud provider dependent. As kubernetes has evolved tremendously, it has become difficult
for different cloudproviders (currently 8), to make changes and iterate quickly. Moreover, the
cloudproviders are constrained by the kubernetes build/release lifecycle. This commit is the first
step in moving towards a kubernetes code base where cloud providers specific code will move out of
the core repository, and will be maintained by the cloud providers themselves.

I have added a new cloud provider called "external", which signals the controller-manager that
cloud provider specific loops are being run by another controller. I have added these changes in such
a way that the existing cloud providers are not affected. This change is completely backwards compatible, and does not require any changes to the way kubernetes is run today.

Finally, along with the controller-manager, the kubelet also has cloud-provider specific code, and that will be addressed in a different commit/issue.

@alena1108 @ibuildthecloud @thockin @dchen1107 

**Special notes for your reviewer**:

@thockin - Im making this **WIP** PR to ensure that I don't stray too far from everyone's view of how we should make this change. As you can see, only one controller, namely `nodecontroller` can be disabled with the `--cloudprovider=external` flag at the moment. I'm working on cleaning up the `rancher-controller-manger` that I wrote to test this.

Secondly, I'd like to use this PR to address cloudprovider specific code in kubelet and api-server.

**Kubelet**
Kubelet uses provider specific code for node registration and for checking node-status. I thought of two ways to divide the kubelet: 
- We could start a cloud provider specific kubelet on each host as a part of kubernetes, and this cloud-specific-kubelet does node registration and node-status checks. 
- Create a kubelet plugin for each provider, which will be started by kubelet as a long running service. This plugin can be packaged as a binary.

I'm leaning towards the first option. That way, kubelet does not have to manage another process, and we can offload the process management of the cloud-provider-specific-kubelet to something like systemd. 

@dchen1107 @thockin what do you think?

**Kube-apiserver**

Kube-apiserver uses provider specific code for distributing ssh keys to all the nodes of a cluster. Do you have any suggestions about how to address this? 

**Release note**:

``` release-note
```

Automatic merge from submit-queue

kubeadm: Default to using token discovery.

Recent changes to support multiple methods for discovery meant that
"kubeadm init" no longer was sufficient and users would need to add
"--discovery token://" to achieve the same results.

Instead lets assume discovery if the user does not specify anything else
to maintain parity and the brevity of our original instructions.


**Release note**:

```release-note
NONE
```

CC @mikedanese @luxas

Automatic merge from submit-queue

Support loading UTF16 files if a byte-order-mark is present

Add support in kubectl for loading UTF16 encoded files if they have a correct BOM (Byte-Order-Mark https://en.wikipedia.org/wiki/Byte_order_mark) at the beginning
of the file. Falls back on UTF8 encoding, if no understandable BOM is present.

Fixes part of https://github.com/kubernetes/kubernetes/issues/39007

@fabianofranz @deads2k @kubernetes/sig-cli-misc

Automatic merge from submit-queue (batch tested with PRs 38920, 38090)

Improve error message for name/label validation.

Instead of just providing regex in name/label validation error output, we need to add the naming rules of the name/label, which is more end-user readable.

Fixed #37654

Automatic merge from submit-queue

Add event when failed to open local port.

fixes #37979 .

Automatic merge from submit-queue

Fix skipping - protobuf fields

**What this PR does / why we need it**: fixes the protobuf generator to skip fields with a protobuf tag of `"-"`

Match changes in https://github.com/kubernetes/gengo/pull/19

I couldn't get godeps to work to vendor this change in from gengo, so I made the same edits manually in cmd/libs/go2idl. A task for another day...

@smarterclayton

Automatic merge from submit-queue (batch tested with PRs 39059, 39175, 35676, 38655)

Fix fsGroup to vSphere

**What this PR does / why we need it**:
Fixes #34039 by adding support for fsGroup to vSphere Volume. 

**Special notes for your reviewer**:
Tested with example from http://stackoverflow.com/questions/35213589/docker-container-with-non-root-user-deployed-in-google-container-engine-can-not
Before this fix got error ```Permission Denied```.

**Release note**:

`NONE`

cc @pdhamdhere @kerneltime @BaluDontu

Automatic merge from submit-queue (batch tested with PRs 39059, 39175, 35676, 38655)

ReplicaSet has onwer ref of the Deployment that created it

**What this PR does / why we need it**:
This enabled garbage collection for ReplicaSets and ensures they are owned by their respective Deployment objects.

fixes https://github.com/kubernetes/kubernetes/issues/33845

This is an initial PR to get feedback. Will update this quickly with unit tests if this seems like in the right direction

Automatic merge from submit-queue (batch tested with PRs 39059, 39175, 35676, 38655)

Revert "Make fluentd pods critical"

Reverts kubernetes/kubernetes#39146

Automatic merge from submit-queue

Don't evict static pods

A follow up to https://github.com/kubernetes/kubernetes/pull/38914, the desired behavior is to:    

1. deprioritize critical pods
2. never evict static pods

I don't think 1 needs a cherrypick if 2 goes in. 

partial fix for #https://github.com/kubernetes/kubernetes/issues/38322

Fixes #39180

Automatic merge from submit-queue (batch tested with PRs 39006, 39078, 37188, 39118)

Cleanup dns

* Remove hostname endpoints annotation (was beta feature)
* Remove references to non-client-go API
* Replaces references to internal kubernetes API with client-go.

Automatic merge from submit-queue (batch tested with PRs 39006, 39078, 37188, 39118)

add test functions in context_test.go

Automatic merge from submit-queue (batch tested with PRs 39006, 39078, 37188, 39118)

Fix bug of missing random watch events in the event stream

Fix #39072

Match changes in https://github.com/kubernetes/gengo/pull/19

Automatic merge from submit-queue

In-cluster configs must take flag overrides into account

**What this PR does / why we need it**: Some flags must override in-cluster configs if provided to `kubectl` inside a cluster.

**Which issue this PR fixes**: Fixes https://github.com/kubernetes/kubernetes/issues/38834

**Release note**:
```release-note
Fixed a bug where the --server, --token, and --certificate-authority flags were not overriding the related in-cluster configs when provided in a `kubectl` call inside a cluster.
```

Addresses: kubernetes/features#88

This commit starts breaking the controller manager into two pieces, namely,

1. cloudprovider dependent piece
2. coudprovider agnostic piece

the controller manager has the following control loops -

   - nodeController
   - volumeController
   - routeController
   - serviceController
   - replicationController
   - endpointController
   - resourcequotacontroller
   - namespacecontroller
   - deploymentController etc..

among the above controller loops,

   - nodeController
   - volumeController
   - routeController
   - serviceController

are cloud provider dependent. As kubernetes has evolved tremendously, it has become difficult
for different cloudproviders (currently 8), to make changes and iterate quickly. Moreover, the
cloudproviders are constrained by the kubernetes build/release lifecycle. This commit is the first
step in moving towards a kubernetes code base where cloud providers specific code will move out of
the core repository, and will be maintained by the cloud providers themselves.

Finally, along with the controller-manager, the kubelet also has cloud-provider specific code, and that will
be addressed in a different commit/issue.

tolerate missing subjectaccessreview in e2e

Replaces references to internal kubernetes API with client-go.