Automatic merge from submit-queue
Bugfix: cloud-controller-manager routeController.run() block in WaitForCacheSync
**What this PR does / why we need it**:
cloud-controller-manager routeController.run not run in goroutine will block to wait nodeListerSynced because sharedInformers not started
**Special notes for your reviewer**:
cloud-controller-manager routeController.run should run in goroutine like the same code in kube-controller-manager,or it will block to wait nodeListerSynced,but sharedInformers not startd.
```go
//controller-manager.go:197
func StartControllers(s *options.CloudControllerManagerServer, kubeconfig *restclient.Config, rootClientBuilder, clientBuilder controller.ControllerClientBuilder, stop <-chan struct{}, recorder record.EventRecorder, cloud cloudprovider.Interface) error {
// Function to build the kube client object
client := func(serviceAccountName string) clientset.Interface {
return rootClientBuilder.ClientOrDie(serviceAccountName)
}
versionedClient := client("shared-informers")
sharedInformers := informers.NewSharedInformerFactory(versionedClient, resyncPeriod(s)())
ocateNodeCIDRs && s.ConfigureCloudRoutes {
if routes, ok := cloud.Routes(); !ok {
glog.Warning("configure-cloud-routes is set, but cloud provider does not support routes. Will not configure cloud provider routes.")
} else {
routeController := routecontroller.New(routes, client("route-controller"), sharedInformers.Core().V1().Nodes(), s.ClusterName, clusterCIDR)
//should run in goroutine
-----> routeController.Run(stop, s.RouteReconciliationPeriod.Duration)
time.Sleep(wait.Jitter(s.ControllerStartInterval.Duration, ControllerStartJitter))
}
} else {
glog.Infof("Will not configure cloud provider routes for allocate-node-cidrs: %v, configure-cloud-routes: %v.", s.AllocateNodeCIDRs, s.ConfigureCloudRoutes)
}
---> sharedInformers.Start(stop)
select {}
}
//routecontroller.go:77
func (rc *RouteController) Run(stopCh <-chan struct{}, syncPeriod time.Duration) {
defer utilruntime.HandleCrash()
glog.Info("Starting the route controller")
//will block
---> if !cache.WaitForCacheSync(stopCh, rc.nodeListerSynced) {
utilruntime.HandleError(fmt.Errorf("timed out waiting for caches to sync"))
return
}
go wait.NonSlidingUntil(func() {
if err := rc.reconcileNodeRoutes(); err != nil {
glog.Errorf("Couldn't reconcile node routes: %v", err)
}
}, syncPeriod, wait.NeverStop)
}
```
Automatic merge from submit-queue
Make the dockershim root directory configurable
Make the dockershim root directory configurable so things like
integration tests (e.g. in OpenShift) can run as non-root.
cc @sttts @derekwaynecarr @yujuhong @Random-Liu @kubernetes/sig-node-pr-reviews @kubernetes/rh-cluster-infra
Automatic merge from submit-queue
Exit from NewController() for PersistentVolumeController when InitPlugins() failed
Exit from NewController() for PersistentVolumeController when InitPlugins() failed just like NewAttachDetachController() does
**Release note**:
```release-note
NONE
```
@jsafrane @saad-ali PTAL. Thanks in advance
Automatic merge from submit-queue
Non global admission plugin registry
For testing the global state is a problem. This PR turns the actual registry into a struct that must be instantiated. For the beginning, we do this in `pkg/kubeapiserver/admission`. In some follow-up (where we hunt down all globals some day), we will move this into the genericapiserver.
Automatic merge from submit-queue
Conversionlint
**What this PR does / why we need it**:
This PR adds a comment placeholder for the exported or public functions in conversion generator files. Such functions without a comment results into golint failure in various generated files. The changes in this patch takes care of about 36 related lint failures.
Given below is an example lint error,
zz_generated.conversion.go:91:1: exported function Convert_v1alpha1_Binding_To_servicecatalog_Binding should have comment or be unexported
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
Considering minor changes no issue is created.
**Special notes for your reviewer**:
Hello, I am trying to fix golint failures as we see them in the kubernetes-incubator/service-catalog project. I have separate PRs opened for lint issues related to other generator code which are in separate libraries like gengo. Thanks!
**Release note**:
```release-note
```
Automatic merge from submit-queue
Edge-based userspace LB in kube-proxy
@thockin @bowei - if one of you could take a look if that PR doesn't break some basic kube-proxy assumptions. The similar change for winuserproxy should be pretty trivial.
And we should also do that for iptables, but that requires splitting the iptables code to syncProxyRules (which from what I know @thockin already started working on so we should probably wait for it to be done).
Automatic merge from submit-queue
Add support for IP aliases for pod IPs (GCP alpha feature)
```release-note
Adds support for allocation of pod IPs via IP aliases.
# Adds KUBE_GCE_ENABLE_IP_ALIASES flag to the cluster up scripts (`kube-{up,down}.sh`).
KUBE_GCE_ENABLE_IP_ALIASES=true will enable allocation of PodCIDR ips
using the ip alias mechanism rather than using routes. This feature is currently
only available on GCE.
## Usage
$ CLUSTER_IP_RANGE=10.100.0.0/16 KUBE_GCE_ENABLE_IP_ALIASES=true bash -x cluster/kube-up.sh
# Adds CloudAllocator to the node CIDR allocator (kubernetes-controller manager).
If CIDRAllocatorType is set to `CloudCIDRAllocator`, then allocation
of CIDR allocation instead is done by the external cloud provider and
the node controller is only responsible for reflecting the allocation
into the node spec.
- Splits off the rangeAllocator from the cidr_allocator.go file.
- Adds cloudCIDRAllocator, which is used when the cloud provider allocates
the CIDR ranges externally. (GCE support only)
- Updates RBAC permission for node controller to include PATCH
```
Automatic merge from submit-queue
Remove alphaProvisioner in PVController and AlphaStorageClassAnnotation
remove alpha annotation and alphaProvisioner
**Release note**:
```release-note
NONE
```
If CIDRAllocatorType is set to `CloudCIDRAllocator`, then allocation
of CIDR allocation instead is done by the external cloud provider and
the node controller is only responsible for reflecting the allocation
into the node spec.
- Splits off the rangeAllocator from the cidr_allocator.go file.
- Adds cloudCIDRAllocator, which is used when the cloud provider allocates
the CIDR ranges externally. (GCE support only)
- Updates RBAC permission for node controller to include PATCH
The exported or public functions without a comment results into golint failure
in various generated files. The changes in this patch takes care of about 36
related lint failures.
Given below is an example lint error,
zz_generated.conversion.go:91:1: exported function
Convert_v1alpha1_Binding_To_servicecatalog_Binding should have comment or be
unexported
Automatic merge from submit-queue (batch tested with PRs 43870, 30302, 42722, 43736)
Admission plugin to merge pod and namespace tolerations for restricting pod placement on nodes
```release-note
This admission plugin checks for tolerations on the pod being admitted and its namespace, and verifies if there is any conflict. If there is no conflict, then it merges the pod's namespace tolerations with the the pod's tolerations and it verifies them against its namespace' whitelist of tolerations and returns. If a namespace does not have its default or whitelist tolerations specified, then cluster level default and whitelist is used. An example of its versioned config:
apiVersion: apiserver.k8s.io/v1alpha1
kind: AdmissionConfiguration
plugins:
- name: "PodTolerationRestriction"
configuration:
apiVersion: podtolerationrestriction.admission.k8s.io/v1alpha1
kind: Configuration
default:
- Key: key1
Value: value1
- Key: key2
Value: value2
whitelist:
- Key: key1
Value: value1
- Key: key2
Value: value2
```
Automatic merge from submit-queue
Add dockershim only mode
This PR added a `experimental-dockershim` hidden flag in kubelet to run dockershim only.
We introduce this flag mainly for cri validation test. In the future we should compile dockershim into another binary.
@yujuhong @feiskyer @xlgao-zju
/cc @kubernetes/sig-node-pr-reviews
Automatic merge from submit-queue
add rancher credential provider
This adds rancher as a credential provider in kubernetes.
@erictune This might be a good opportunity to discuss adding a provision for people to have their own credential providers that is similar to the new cloud provider changes (https://github.com/kubernetes/community/pull/128). WDYT?
```
release-note
Added Rancher Credential Provider to use Rancher Registry credentials when running in a Rancher cluster
```
Automatic merge from submit-queue (batch tested with PRs 43777, 44121)
Add patchMergeKey and patchStrategy support to OpenAPI
Support generating Open API extensions for strategic merge patch tags in go struct tags
Support `patchStrategy` and `patchMergeKey`.
Also support checking if the Open API extension and struct tags match.
```release-note
Support generating Open API extensions for strategic merge patch tags in go struct tags
```
cc: @pwittrock @ymqytw
(Description mostly copied from #43833)
Automatic merge from submit-queue (batch tested with PRs 43951, 43386)
kubeadm: Fix issue when kubeadm reset isn't working and the docker service is disabled
**What this PR does / why we need it**:
If the docker service is disabled, the preflight check lib will return a warning.
That warning _should not_ matter when deciding whether to reset docker state or not.
The current code skips the docker reset if the docker service is disabled, which is a bug.
Also, `Check()` must not return a `nil` slice.
It should be added that I **really don't like what we have at the moment**, I'd love to discuss with the node team to add something to CRI that basically says, "remove everything on this node" so we can stop doing this. Basically, kubeadm could talk to the specified socket (by default dockershim.sock), and call the CRI interface and say that everything should be cleaned up. This would then be cross-CRI-implementation at the same time and would work if you're using rkt, cri-o or whatever.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*:
helps in https://github.com/kubernetes/kubernetes/issues/43950
**Special notes for your reviewer**:
**Release note**:
```release-note
kubeadm: Make `kubeadm reset` tolerant of a disabled docker service.
```
@mikedanese @jbeda @dmmcquay @pipejakob @yujuhong @freehan
Automatic merge from submit-queue
Adding krousey as a kubeadm reviewer and owner
I would like to join the illustrious ranks of kubeadm owners. I plan to spend a considerable amount of time integrating this tool into our GCE and GKE deployments. If approver is too much, I would still like to be a reviewer.
I will mark this as "Do not merge" until I see approval from all current owners.
Automatic merge from submit-queue (batch tested with PRs 44143, 44133)
Fix panic in kubeadm master node setup
The problem was [caught](https://travis-ci.org/Mirantis/kubeadm-dind-cluster/jobs/218999640#L3249) by kubeadm-dind-cluster CI.
```
[kubeadm] WARNING: kubeadm is in beta, please do not use it for production clusters.
[init] Using Kubernetes version: v1.6.1
[init] Using Authorization mode: RBAC
[preflight] Skipping pre-flight checks
[certificates] Generated CA certificate and key.
[certificates] Generated API server certificate and key.
[certificates] API Server serving cert is signed for DNS names [kube-master kubernetes kubernetes.default kubernetes.default.svc kubernetes.default.svc.cluster.local] and IPs [10.96.0.1 10.192.0.2]
[certificates] Generated API server kubelet client certificate and key.
[certificates] Generated service account token signing key and public key.
[certificates] Generated front-proxy CA certificate and key.
[certificates] Generated front-proxy client certificate and key.
[certificates] Valid certificates and keys now exist in "/etc/kubernetes/pki"
[kubeconfig] Wrote KubeConfig file to disk: "/etc/kubernetes/scheduler.conf"
[kubeconfig] Wrote KubeConfig file to disk: "/etc/kubernetes/admin.conf"
[kubeconfig] Wrote KubeConfig file to disk: "/etc/kubernetes/kubelet.conf"
[kubeconfig] Wrote KubeConfig file to disk: "/etc/kubernetes/controller-manager.conf"
[apiclient] Created API client, waiting for the control plane to become ready
[apiclient] All control plane components are healthy after 19.017839 seconds
panic: assignment to entry in nil map
goroutine 1 [running]:
panic(0x1b62140, 0xc4203f0380)
/usr/local/go/src/runtime/panic.go:500 +0x1a1
k8s.io/kubernetes/cmd/kubeadm/app/phases/apiconfig.attemptToUpdateMasterRoleLabelsAndTaints(0xc420b18be0, 0x4e, 0x0)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/cmd/kubeadm/app/phases/apiconfig/setupmaster.go:57 +0x15b
k8s.io/kubernetes/cmd/kubeadm/app/phases/apiconfig.UpdateMasterRoleLabelsAndTaints(0xc420b18be0, 0x1a, 0xc420b18be0)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/cmd/kubeadm/app/phases/apiconfig/setupmaster.go:86 +0x2f
k8s.io/kubernetes/cmd/kubeadm/app/cmd.(*Init).Run(0xc4201a4040, 0x29886e0, 0xc420022010, 0x1c73d01, 0xc4201a4040)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/cmd/kubeadm/app/cmd/init.go:220 +0x29c
k8s.io/kubernetes/cmd/kubeadm/app/cmd.NewCmdInit.func1(0xc4203a46c0, 0xc420660680, 0x0, 0x2)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/cmd/kubeadm/app/cmd/init.go:86 +0x197
k8s.io/kubernetes/vendor/github.com/spf13/cobra.(*Command).execute(0xc4203a46c0, 0xc420660560, 0x2, 0x2, 0xc4203a46c0, 0xc420660560)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/vendor/github.com/spf13/cobra/command.go:603 +0x439
k8s.io/kubernetes/vendor/github.com/spf13/cobra.(*Command).ExecuteC(0xc4203b1d40, 0xc4203a4b40, 0xc4203a46c0, 0xc4203a4000)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/vendor/github.com/spf13/cobra/command.go:689 +0x367
k8s.io/kubernetes/vendor/github.com/spf13/cobra.(*Command).Execute(0xc4203b1d40, 0xc42046c420, 0x29886a0)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/vendor/github.com/spf13/cobra/command.go:648 +0x2b
k8s.io/kubernetes/cmd/kubeadm/app.Run(0xc420627f70, 0xc4200001a0)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/cmd/kubeadm/app/kubeadm.go:35 +0xe8
main.main()
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/cmd/kubeadm/kubeadm.go:26 +0x22
```
Automatic merge from submit-queue
[Federation] Remove FEDERATIONS_DOMAIN_MAP references
Remove all references to FEDERATIONS_DOMAIN_MAP as this method is no longer is used and is replaced by adding federation domain map to kube-dns configmap.
cc @madhusudancs @kubernetes/sig-federation-pr-reviews
**Release note**:
```
[Federation] Mechanism of adding `federation domain maps` to kube-dns deployment via `--federations` flag is superseded by adding/updating `federations` key in `kube-system/kube-dns` configmap. If user is using kubefed tool to join cluster federation, adding federation domain maps to kube-dns is already taken care by `kubefed join` and does not need further action.
```
Automatic merge from submit-queue (batch tested with PRs 44097, 42772, 43880, 44031, 44066)
kubeadm: Wait for node before updating labels and taints
**What this PR does / why we need it**:
Adds again (removed in #43881) waiting for at last single node appearance during kubeadm attempt to update master role labels and taints.
**Which issue this PR fixes**:
fixeskubernetes/kubeadm#221
**Release note**:
```NONE
```
Use shared informers instead of creating local controllers/reflectors
for the proxy's endpoints and service configs. This allows downstream
integrators to pass in preexisting shared informers to save on memory &
cpu usage.
This also enables the cache mutation detector for kube-proxy for those
presubmit jobs that already turn it on.
Kubelet flags are not necessarily appropriate for the KubeletConfiguration
object. For example, this PR also removes HostnameOverride and NodeIP
from KubeletConfiguration. This is a preleminary step to enabling Nodes
to share configurations, as part of the dynamic Kubelet configuration
feature (#29459). Fields that must be unique for each node inhibit
sharing, because their values, by definition, cannot be shared.
Automatic merge from submit-queue
kubelet: change image-gc-high-threshold below docker dm.min_free_space
docker dm.min_free_space defaults to 10%, which "specifies the min free space percent in a thin pool require for new device creation to succeed....Whenever a new a thin pool device is created (during docker pull or during container creation), the Engine checks if the minimum free space is available. If sufficient space is unavailable, then device creation fails and any relevant docker operation fails." [1]
This setting is preventing the storage usage to cross the 90% limit. However, image GC is expected to kick in only beyond image-gc-high-threshold. The image-gc-high-threshold has a default value of 90%, and hence GC never triggers. If image-gc-high-threshold is set to a value lower than (100 - dm.min_free_space)%, GC triggers.
xref https://bugzilla.redhat.com/show_bug.cgi?id=1408309
```release-note
changed kubelet default image-gc-high-threshold to 85% to resolve a conflict with default settings in docker that prevented image garbage collection from resolving low disk space situations when using devicemapper storage.
```
@derekwaynecarr @sdodson @rhvgoyal
Automatic merge from submit-queue
fix typo in kubeadm join -h
```
Flags:
--config string Path to kubeadm config file
--discovery-file string A file or url from which to load cluster information
--discovery-token string A token used to validate cluster information fetched from the master
--skip-preflight-checks skip preflight checks normally run before modifying the system
--tls-bootstrap-token string A token used for TLS bootstrapping
--token string Use this token for both discovery-token and tls-bootstrap-token
```
Automatic merge from submit-queue
Make RBAC post-start hook conditional on RBAC authorizer being used
Makes the RBAC post-start hook (and reconciliation) conditional on the RBAC authorizer being used
Ensures we don't set up unnecessary objects.
```release-note
RBAC role and rolebinding auto-reconciliation is now performed only when the RBAC authorization mode is enabled.
```
Automatic merge from submit-queue (batch tested with PRs 42360, 43109, 43737, 43853)
kubeadm: test-cmds for kubeadm completion
**What this PR does / why we need it**: Adding test-cmds for kubeadm completion.
Adding tests is a WIP from #34136
**Special notes for your reviewer**: /cc @luxas
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue
don't wait for first kubelet to be ready and drop dummy deploy
Per https://github.com/kubernetes/kubernetes/issues/43815#issuecomment-290270198, I suggest that we drop both the node ready and the dummy deployment check altogether for 1.6 and move them to a validation phase for 1.7.
I really think we should drop these checks altogether. CreateClientAndWaitForAPI should create a client and wait for the API, not create dummy deployments and wait for nodes to register and be healthy. These are end to end validations and this is the wrong place to do this stuff. We need an explicit final validation phase for this.
```release-note
Fix a deadlock in kubeadm master initialization.
```
Fixes#43815
Automatic merge from submit-queue (batch tested with PRs 42617, 43247, 43509, 43644, 43820)
Bugfix: OpenAPI-gen was not generating extensions correctly
Fixes a bug in openapi-gen that generated invalid code if x-kubernetes extensions defined in types.go. The location of VendorExtensions was wrong.
Automatic merge from submit-queue (batch tested with PRs 42835, 42974)
remove legacy insecure port options from genericapiserver
The insecure port has been a source of problems and it will prevent proper aggregation into a cluster, so the genericapiserver has no need for it. In addition, there's no reason for it to be in the main kube-apiserver flow either. This pull removes it from genericapiserver and removes it from the shared kube-apiserver code. It's still wired up in the command, but its no longer possible for someone to mess up and start using in mainline code.
@kubernetes/sig-api-machinery-misc @ncdc
Automatic merge from submit-queue
proxy to IP instead of name, but still use host verification
I think I found a setting that lets us proxy to an IP and still do hostname verification on the certificate.
@liggitt @sttts Can you see if you agree that this knob does what I think it does? Last commit only, still needs tests.
Automatic merge from submit-queue (batch tested with PRs 42900, 43044, 42896, 43308, 43621)
require codecfactory
The genericapiserver requires a codec to start. Help new comers to the API by forcing them to set it when they create a new config.
Automatic merge from submit-queue (batch tested with PRs 43149, 41399, 43154, 43569, 42507)
kubeadm: only print stderr/stdout if failed test
**What this PR does / why we need it**: This PR changes when stdout/stderr will be logged during a kubeadm test-cmd test. It's useful when a real failure occurs to only see the failure rather than output that looks like it might be a failure
**Special notes for your reviewer**: /cc @luxas @marun
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue
break kube-apiserver start into stages
This is a code shuffle which breaks the kube-apiserver start into
1. set defaults on the options
1. create the generic config from the options
1. create the master config from the generic config and the options
This makes apiserver composition easy/possible later on.
Automatic merge from submit-queue (batch tested with PRs 43144, 42671, 43226, 43314, 43361)
don't start controllers against unhealthy master
Operating against an unhealthy apiserver is unpredictable. Some clients like `kubectl` need to be best effort in this regard so that you can debug broken apiservers. Controllers shouldn't run against unhealthy masters.
Automatic merge from submit-queue (batch tested with PRs 43144, 42671, 43226, 43314, 43361)
start informers as a post-start-hook
Switches the shared informer start to a post start hook to make future API server composition easier. PostStartHooks will have to be unioned for server composition and this ensures that we don't accidentally skip starting them.
Automatic merge from submit-queue
Better messaging when GKE certificate signing fails.
**What this PR does / why we need it**:
On errors, the GKE signing API can respond with a JSON body that contains an error message explaining the failure. If we're able to extract it, use that message when reporting the error instead of the generic error returned by the webhook library. Also, always add an event to the CSR object on signing errors.
**Release note**:
```release-note
NONE
```
CC @mikedanese @jcbsmpsn
Automatic merge from submit-queue (batch tested with PRs 43642, 43170, 41813, 42170, 41581)
Enable storage class support in Azure File volume
**What this PR does / why we need it**:
Support StorageClass in Azure file volume
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
Support StorageClass in Azure file volume
```
Automatic merge from submit-queue
kubeadm: Remove an outdated comment
Now that `AdvertiseAddress` is a `string` and not
`AdvertiseAddresses` a `[]string` this comment is no longer
necessary.
@k8s-mirror-cluster-lifecycle-misc RFR
**What this PR does / why we need it**
Just a little house cleaning by removing an outdated comment.
**Release note**:
```release-note
NONE
```
On errors, the GKE signing API can respond with a JSON body that
contains an error message explaining the failure. If we're able to
extract it, use that message when reporting the error instead of the
generic error returned by the webhook library. Also, always add an event
to the CSR object on signing errors.
Kubernetes Submit Queue