Automatic merge from submit-queue (batch tested with PRs 58517, 57642). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Fix event message when processing loadbalancer update
**What this PR does / why we need it**:
When a service get updated, in func [processServiceUpdate](https://github.com/kubernetes/kubernetes/blob/master/pkg/controller/service/service_controller.go#L249), we process its LB accordingly, that is, create one if the service requests and no corresponding loadbalancer exists; and delete potential orphaned load balancer if the service does not need it any more.
But if a service does not `wantsLoadBalancer` but get error when trying to `GetLoadBalancer`, user could find an event in format of "CreatingLoadBalancerFailed..."[here](https://github.com/kubernetes/kubernetes/blob/master/pkg/controller/service/service_controller.go#L261), which would confusing users. So we should generate event info according to service type.
**Special notes for your reviewer**:
/sig network
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 57868, 58284, 56370, 58400, 58439). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Refactor service controller to common controller pattern
**What this PR does / why we need it**:
The service controller currently uses a non-standard controller pattern that adds unneeded complexity. This PR moves the service controller to use common tools like the rate limited queue with exponential backoff versus tracking retry backoff in the cachedService object.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes#55528
**Special notes for your reviewer**:
This does change the retry backoff from a linear backoff to an exponential backoff. The min and max retry delay stays the same.
There was only one case where we didn't want to retry, which was when we updated the service status since that handled retries itself. In that case I moved to the common pattern of using `runtime.HandleError`.
**Release note**:
```release-note
NONE
```
/cc @luxas @wlan0 @andrewsykim @thockin
@prydie
/sig cluster-lifecycle
/sig networking
/kind cleanup
Automatic merge from submit-queue (batch tested with PRs 58375, 58306, 58386, 57420, 58035). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Track run status explicitly rather than non-nil check on stopCh
Fixes#57044
GC and quota controllers use a non-nil stop channel as a signal Run() has been called, so ensure that condition holds even when a nil stop channel is passed in
```release-note
NONE
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
[quota controller] remove extra queue.Add()
requeue immediately after an error may end-up with hot-loop
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
[Deployments] Move some tests to use go sub-test
**What this PR does / why we need it**:
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
Fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 57266, 58187, 58186, 46245, 56509). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Avoid panic in Cloud CIDR Allocator
**What this PR does / why we need it**:
I suspect a race exists where we attempt to look up the CIDR for a terminating node. By the time `updateCIDRAllocation` is called the node has disappeared. We determine it does not have a cloud CIDR (i.e. Alias IP Range) and attempt to record a `CIDRNotAvailable` node status. Unfortunately we reference `node.Name` while `node` is still nil.
By getting the node before looking up the cloud CIDR we avoid the nil pointer dereference, and potentially fail fast in the case the node has disappeared.
**Which issue(s) this PR fixes**:
Fixes#58181
**Release note**:
```release-note
Avoid panic when failing to allocate a Cloud CIDR (aka GCE Alias IP Range).
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Comment typo fix in StatefulSet controller
**What this PR does / why we need it**:
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
Fixes #
**Special notes for your reviewer**:
**Release note**:
```
NONE
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
use shared informers for BootstrapSigner controller
**What this PR does / why we need it**:
fix TODO: Switch to shared informers
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
Fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
remove duplicate function
remove duplicate function
**What this PR does / why we need it**:
remove duplicate function `GetDeploymentConditionInternal`
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
Fixes #
**Special notes for your reviewer**:
**Release note**:
```
NONE
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Revert "no need delete endpoint explicitly in endpoint controller"
Reverts kubernetes/kubernetes#57747
For:
1. `checkLeftoverEndpoints` does not work any more.
2. service delete auto trigger endpoint delete in apiserver, if service is successfully deleted but fail on endpoint delete. Then this endpoint will be garbage resource.
The resource quota controller worker pool can deadlock when:
* Worker goroutines are idle waiting for work from queues
* The Sync() method detects discovery updates to apply
The problem is workers acquire a read lock while idle, making write lock
acquisition dependent upon the presence of work in the queues.
The Sync() method blocks on a pending write lock acquisition and won't unblock
until every existing worker processes one item from their queue and releases
their read lock. While the Sync() method's lock is pending, all new read lock
acquisitions will block; if a worker does process work and release its lock, it
will then become blocked on a read lock acquisition; they become blocked on
Sync(). This can easily deadlock all the workers processing from one queue while
any workers on the other queue remain blocked waiting for work.
Fix the deadlock by refactoring workers to acquire a read lock *after* work is
popped from the queue. This allows writers to get locks while workers are idle,
while preserving the worker pause semantics necessary to allow safe sync.
Automatic merge from submit-queue (batch tested with PRs 55316, 57934). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Increase test coverage in `pkg/controllers/podautoscaler/metrics`
**What this PR does / why we need it**:
Increase test coverage for `pkg/controllers/podautoscaler/metrics`. The biggest addition is unit tests for `pkg/controller/podautoscaler/metrics/utilization.go`.
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 55316, 57934). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Make StatefulSet report an event when recreating failed pod
When StatefulSet finds any of its pods in phase Failed it will delete it and create it again. While doing so it will emit delete+create event which is really confusing for the user as he has no idea what's happening.
Fixes https://github.com/kubernetes/kubernetes/issues/58082
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
fix populateDesiredStateOfWorld bug for attach/detach controller
**What this PR does / why we need it**:
The `attachDetachController.populateDesiredStateOfWorld` mehod will fetch all pods from the informer cache and add each pods to `DesiredStateOfWorld` by the [addPod method](da9a4d5da9/pkg/controller/volume/attachdetach/attach_detach_controller.go (L338)).
However, the element type of pod list returned by `podLister.List` is already `*v1.Pod`, so we needn't to take the address any more.
**Release note**:
```release-note
NONE
```
/sig storage
/kind bug
Automatic merge from submit-queue (batch tested with PRs 57926, 57930). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
use sets.String to replace slice when sort []string
**What this PR does / why we need it**:
use sets.String to replace slice when sort []string
No need to maintain a slice comparison function.
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
Fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Move scheduler out of plugin directory
**What this PR does / why we need it**:
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
This is but one step toward resolving the referenced issue.
/ref #57579
**Special notes for your reviewer**:
**Release note**:
```release-note
Default scheduler code is moved out of the plugin directory.
plugin/pkg/scheduler -> pkg/scheduler
plugin/cmd/kube-scheduler -> cmd/kube-scheduler
```
/sig scheduling
This moves plugin/pkg/scheduler to pkg/scheduler and
plugin/cmd/kube-scheduler to cmd/kube-scheduler.
Bulk of the work was done with gomvpkg, except for kube-scheduler main
package.
Automatic merge from submit-queue (batch tested with PRs 57900, 57888). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Re-add nodecontroller OWNERS file
Re-add OWNERS file that got moved to nodeipam directory.
cc @cheftako
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
optimize volumeResizeMap lock
**What this PR does / why we need it**:
volumeResizeMap implements VolumeResizeMap interface, but has several useless lock section and one too large critical section.
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
Fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Split the NodeController into lifecycle and ipam pieces.
**What this PR does / why we need it**: Separates node controller into ipam and lifecycle components.
Prepatory work for removing cloud provider dependency from node
controller running in Kube Controller Manager. Splitting the node
controller into its two major pieces life-cycle and CIDR/IP
management. Both pieces currently need the the cloud system to do their work.
Removing lifecycles dependency on cloud will be fixed ina followup PR.
Moved node scheduler code to live with node lifecycle controller.
Got the IPAM/Lifecycle split completed. Still need to rename pieces.
Made changes to the utils and tests so they would be in the appropriate
package.
Moved the node based ipam code to nodeipam.
Made the relevant tests pass.
Moved common node controller util code to nodeutil.
Removed unneeded pod informer sync from node ipam controller.
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
Fixes #https://github.com/kubernetes/kubernetes/issues/52369
**Special notes for your reviewer**:
**Release note**:
```release-note
None
```
Automatic merge from submit-queue (batch tested with PRs 57810, 57752). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Clarify error messages in HPA metrics
**What this PR does / why we need it**:
With the introduction of the RESTMetrics client, there are two ways to
fetch metrics for auto-scaling. However, they previously shared error
messages. This could be misleading. Make the error message more clearly
show which method is in use.
**Special notes for your reviewer**:
@DirectXMan12 re conversation in #56827
**Release note**:
```release-note
NONE
```
Prepatory work fpr removing cloud provider dependency from node
controller running in Kube Controller Manager. Splitting the node
controller into its two major pieces life-cycle and CIDR/IP
management. Both pieces currently need the the cloud system to do their work.
Removing lifecycles dependency on cloud will be fixed ina followup PR.
Moved node scheduler code to live with node lifecycle controller.
Got the IPAM/Lifecycle split completed. Still need to rename pieces.
Made changes to the utils and tests so they would be in the appropriate
package.
Moved the node based ipam code to nodeipam.
Made the relevant tests pass.
Moved common node controller util code to nodeutil.
Removed unneeded pod informer sync from node ipam controller.
Fixed linter issues.
Factored in feedback from @gmarek.
Factored in feedback from @mtaufen.
Undoing unneeded change.
With the introduction of the RESTMetrics client, there are two ways to
fetch metrics for auto-scaling. However, they previously shared error
messages. This could be misleading. Make the error message more clearly
show which method is in use.
Automatic merge from submit-queue (batch tested with PRs 49856, 56257, 57027, 57695, 57432). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
Change to pkg/util/node.UpdateNodeStatus
**What this PR does / why we need it**:
> // TODO: Change to pkg/util/node.UpdateNodeStatus.
**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**:
/cc @brendandburns @dchen1107 @lavalamp
**Release note**:
```release-note
None
```
Automatic merge from submit-queue (batch tested with PRs 57747, 57749). If you want to cherry-pick this change to another branch, please follow the instructions <a href="https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md">here</a>.
no need delete endpoint explicitly in endpoint controller
**What this PR does / why we need it**:
It is no need to do delete in controller. Because in service restStorage, endpoint is deleted when delete service. See https://github.com/kubernetes/kubernetes/blob/master/pkg/registry/core/service/rest.go#L170~L180
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
Fixes#57745
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Kubernetes Submit Queue