Automatic merge from submit-queue. If you want to cherry-pick this change to another branch, please follow the instructions here: https://github.com/kubernetes/community/blob/master/contributors/devel/cherry-picks.md.
Change CPU sample sanitization in HPA.
**What this PR does / why we need it**:
Change CPU sample sanitization in HPA.
Ignore samples if:
- Pod is beeing initalized - 5 minutes from start defined by flag
- pod is unready
- pod is ready but full window of metric hasn't been colected since
transition
- Pod is initialized - 5 minutes from start defined by flag:
- Pod has never been ready after initial readiness period.
**Release notes:**
```release-note
Improve CPU sample sanitization in HPA by taking metric's freshness into account.
```
Ignore samples if:
- Pod is beeing initalized - 5 minutes from start defined by flag
- pod is unready
- pod is ready but full window of metric hasn't been colected since
transition
- Pod is initialized - 5 minutes from start defined by flag:
- Pod has never been ready after initial readiness period.
This is the old behaviour and we did not intent to change it due to enabled authn/z in general.
As the kube-apiserver this sets the "system:unsecured" user info.
Automatic merge from submit-queue (batch tested with PRs 66916, 67252, 67794, 67619, 67328). 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 HPA sample sanitization
**What this PR does / why we need it**: @mwielgus pointed out a case when HPA fails as a result of my changes to HPA algorithm:
- Have pods that use a lot of CPU during initilization, become ready right after they initialize,
- Trigger a scale up,
- When new pods become ready will will count their usage (even though it's not related to any work that needs doing),
- This triggers another scale up, even though existing pods can handle work, no problem.
The fix is:
- Use all samples for non-cpu metrics.
- Only use CPU samples if:
- Pod is ready and was started more than 2 minutes ago, or
- Pod is unready and last readiness change happened more than 10s after it was started.
Reasoning behind this in: https://docs.google.com/document/d/1UdtYedhmCxjaJIQi6hwJMY0eHQQKxlVD8lSHZC1BPOA/edit
**Which issue(s) this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close the issue(s) when PR gets merged)*:
**Special notes for your reviewer**:
**Release note**:
```release-note
Replace scale up forbidden window with disregarding CPU samples collected when pod was initializing.
```
Duration of initialization taint on CPU and window of initial readiness
setting controlled by flags.
Adding API violation exceptions following example of e50340ee23
Automatic merge from submit-queue (batch tested with PRs 62441, 66702, 67254, 67421, 65309). 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>.
Add backoff for DS's pod deletion to limit fighting with kubelet failing the pod repeatedly
**What this PR does / why we need it**:
Limits consequences of DS controller on hot loop fighting with kubelet.
**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/65240
**Release note**:
```release-note
DaemonSet controller is now using backoff algorithm to avoid hot loops fighting with kubelet on pod recreation when a particular DaemonSet is misconfigured.
```
TODO:
- [x] Export the backoff settings as args or constants
- [x] Add test a case
/cc @mfojtik
(Will add more folks when it's ready, to avoid spamming them.)
Automatic merge from submit-queue (batch tested with PRs 67347, 67307, 67358, 67364, 67385). 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>.
*-controller-manager: remove the redundancy import file
**What this PR does / why we need it**:
remove the redundancy import file
**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**:
remove the redundancy import file @deads2k @mikedanese
**Release note**:
```release-note
NONE
```
Instead discard metric values for pods that are unready and have never
been ready (they may report misleading values, the original reason for
introducing scale up forbidden window).
Use per pod metric when pod is:
- Ready, or
- Not ready but creation timestamp and last readiness change are more
than 10s apart.
In the latter case we asume the pod was ready but later became unready.
We want to use metrics for such pods because sometimes such pods are
unready because they were getting too much load.
Automatic merge from submit-queue (batch tested with PRs 64664, 65836, 65917). 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 stray comment from a merge
Signed-off-by: Davanum Srinivas <davanum@gmail.com>
**What this PR does / why we need it**:
Remove stray comment from a merge.
**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 65456, 65549). 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 flexvolume in containerized kubelets
Fixes flex volumes in containerized kubelets.
cc @jsafrane @chakri-nelluri @verult
Note to reviewers : e2e tests pass in local containarized cluster.
```release-note
Fix flexvolume in containarized kubelets
```
Automatic merge from submit-queue (batch tested with PRs 65897, 65909, 65856, 65815). 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>.
only need to ignore resources that match discovery conditions
GC and quota controllers ignore resources that are too expensive to manage. In kube this is only events. The incompatible resources should now be excluded on the basis of discovery. We should actually reflect that in the RESTStorage (done for GC for events) and discovery too.
@liggitt
@kubernetes/sig-api-machinery-bugs
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 64511, 65865, 65867, 65866). 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>.
kube-controller-manager: add stopCh plumbing
Needed for integration testing.
Automatic merge from submit-queue (batch tested with PRs 65094, 65533, 63522, 65694, 65702). 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>.
set leader election client and renew timeout
**What this PR does / why we need it**:
set leader-election client timeout
set timeout for tryAcquireOrRenew
**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#65090#65257
**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>.
Add missing error handling in schema-related code
**What this PR does / why we need it**:
Adds missing error handling to a few places.
**Which issue(s) this PR fixes**
Updates #51457. Still more work to do to fix the issue - client generation code needs to be updated (addressed in https://github.com/kubernetes/kubernetes/pull/64664).
**Release note**:
```release-note
NONE
```
/kind bug
/sig api-machinery
Automatic merge from submit-queue (batch tested with PRs 65256, 64236, 64919, 64879, 57932). 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>.
Cancellable leader election
**What this PR does / why we need it**:
Adds ability to cancel leader election. Useful in integration tests where the whole app is started and stopped in each test.
**Special notes for your reviewer**:
I used the `context` package - it is impossible/hard to achieve the same behaviour with just channels without spawning additional goroutines but it is trivial with `context`. See `acquire()` and `renew()` methods.
**Release note**:
```release-note
NONE
```
/kind enhancement
/sig api-machinery
Automatic merge from submit-queue (batch tested with PRs 57082, 64325, 64016, 64443, 64403). 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>.
should not ignore err when convert api version
**What this PR does / why we need it**:
should not ignore err when convert api version
**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 62756, 63862, 61419, 64015, 64063). 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 apps/v1 Deployment/ReplicaSet in controller and kubectl
This updates the Deployment controller and integration/e2e tests to use apps/v1, as part of #55714.
This also requires updating any other components that use the `deployment/util` package, most notably `kubectl`. That means client versions 1.11 and above will only work with server versions 1.9 and above. This is well within our client-server version skew policy of +/-1 minor version.
However, this PR *only* updates the parts of `kubectl` that used `deployment/util`. So although kubectl now requires apps/v1, it still also depends on extensions/v1beta1. Migrating other parts of kubectl to apps/v1 is beyond the scope of this PR, which was just to change the Deployment controller and fix all the fallout.
```release-note
kubectl: This client version requires the `apps/v1` APIs, so it will not work against a cluster version older than v1.9.0. Note that kubectl only guarantees compatibility with clusters that are +/-1 minor version away.
```
Automatic merge from submit-queue (batch tested with PRs 63049, 59731). 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-enable nodeipam in kube-controller-manager
**What this PR does / why we need it**:
Re-enables nodeipam controller for external clouds. Also does a small refactor so that we don't need to pass in `allocateNodeCidr` into the controller.
In v1.10 we made a change (9187b343e1 (diff-f11913dc67d80d36b3d06a93f61c49cf) in https://github.com/kubernetes/kubernetes/pull/57492) where nodeipam would be disabled for any cluster that sets `--cloud-provider=external`. The original intention behind this was that the nodeipam controller is cloud specific for some clouds (only GCE at the moment) so it should be moved to the CCM (cloud controller manager). After some discussions with wg-cloud-provider it makes sense to re-enable nodeipam controller in KCM and have GCE CCM enable its own cloud-specific IPAM controller as part of [Initialize()](https://github.com/kubernetes/kubernetes/blob/master/pkg/cloudprovider/cloud.go#L33-L35). This would allow for GCE to run nodeipam in both KCM (by setting --cloud-provider=gce and --allocate-node-cidr) and in the CCM (once implemented in `Initialize()`) without disabling nodeipam in the KCM for all external clouds and avoids having to implement nodeipam in CCM.
**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
Re-enable nodeipam controller for external clouds.
```
Automatic merge from submit-queue (batch tested with PRs 63526, 60371, 63444). 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>.
update garbage collection to use the new dynamic client
Update GC to use the new and easy to use dynamic client. This is one of two remaining stragglers.
@kubernetes/sig-api-machinery-pr-reviews
@caesarxuchao @ironcladlou
```release-note
NONE
```
https://github.com/kubernetes/kubernetes/pull/62913 switched from using a client pool, where each groupVersionResource got its own rest client, to a single client.
This increases the QPS to account for increased requests using a single rest client rate limiter.
After K8s 1.10 is upgraded to K8s 1.11 finalizer [kubernetes.io/pvc-protection] is added to PVCs
because StorageObjectInUseProtection feature will be GA in K8s 1.11.
However, when K8s 1.11 is downgraded to K8s 1.10 and the StorageObjectInUseProtection feature is disabled
the finalizers remain in the PVCs and as pvc-protection-controller is not started in K8s 1.10 finalizers
are not removed automatically from deleted PVCs and that's why deleted PVC are not removed from the system
but remain in Terminating phase.
The same applies to pv-protection-controller and [kubernetes.io/pvc-protection] finalizer in PVs.
That's why pvc-protection-controller is always started because the pvc-protection-controller removes finalizers
from PVCs automatically when a PVC is not in active use by a pod.
Also the pv-protection-controller is always started to remove finalizers from PVs automatically when a PV is not
Bound to a PVC.
Related issue: https://github.com/kubernetes/kubernetes/issues/60764
Automatic merge from submit-queue (batch tested with PRs 61306, 60270, 62496, 62181, 62234). 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 up the huge set of flags into smaller option structs
**What this PR does / why we need it**:
To make generic, we do following work:
1. Spliting `KubeControllerManagerConfiguration` in kube-controller-manager and cloud-controller-manager into fewer smaller struct options order by controller, and modify relative flag. Also part of #59483.
2. Spliting `componentconfig` in controller-manager into fewer smaller config order by controller too.
All works follow #59582, using `option+config` logic.
**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
```
Thanks to some great sleuthing by ikruglov!
kube-controller-manager defaults --leader-elect to true. We should
do the same for kube-scheduler. kube-scheduler used to have this
set to true, but it got lost during refactoring in:
efb2bb71cd
Automatic merge from submit-queue (batch tested with PRs 60455, 61365, 61375, 61597, 61491). 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>.
Check apps/v1 StatefulSet available before starting its controller
**What this PR does / why we need it**: StatefulSet controller was bumped to use `apps/v1.StatefulSet` already. Without this change, StatefulSet controller will continue to work, but will be broken when `apps/v1beta2.StatefulSet` is removed or disabled.
**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 60376, 55584, 60358, 54631, 60291). 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>.
cloud-controller-manager get /healthz to wait for apiserver to be healthy
**What this PR does / why we need it**:
currently cloud-controller-manager use `restclient.ServerAPIVersions()` to wait for apiserver to be healthy.
Remove ServerAPIVersions and make use of /healthz as all other components do.
**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#60288
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 60054, 60202, 60219, 58090, 60275). 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>.
Fixes for HTTP/2 max streams per connection setting
**What this PR does / why we need it**:
This PR makes two changes. One is to introduce a parameter
for the HTTP/2 setting that an api-server sends to its clients
telling them how many streams they may have concurrently open in
an HTTP/2 connection. If left at its default value of zero,
this means to use the default in golang's HTTP/2 code (which
is currently 250; see https://github.com/golang/net/blob/master/http2/server.go).
The other change is to make the recommended options for an aggregated
api-server set this limit to 1000. The limit of 250 is annoyingly low
for the use case of many controllers watching objects of Kinds served
by an aggregated api-server reached through the main api-server (in
its mode as a proxy for the aggregated api-server, in which it uses a
single HTTP/2 connection for all calls proxied to that aggregated
api-server).
**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#60042
**Special notes for your reviewer**:
**Release note**:
```release-note
Introduced `--http2-max-streams-per-connection` command line flag on api-servers and set default to 1000 for aggregated API servers.
```
Automatic merge from submit-queue (batch tested with PRs 59286, 59743, 59883, 60190, 60165). 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>.
DaemonSet Controller and tests to apps/v1
**What this PR does / why we need it**:
Updates the DaemonSet controller, its integration tests, and its e2e tests to use the apps/v1 API.
**Release note**:
```release-note
The DaemonSet controller, its integration tests, and its e2e tests, have been updated to use the apps/v1 API.
```
Automatic merge from submit-queue (batch tested with PRs 59463, 59719, 60181, 58283, 59966). 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>.
components pprof profiling make use of existing genericapiserver's
**What this PR does / why we need it**:
fix#60278
Instead of writing private pprof, all components make use of generic apiserver existing profiling.
**Release note**:
```release-note
NONE
```
This PR makes two changes. One is to introduce a parameter
for the HTTP/2 setting that an api-server sends to its clients
telling them how many streams they may have concurrently open in
an HTTP/2 connection. If left at its default value of zero,
this means to use the default in golang's HTTP/2 code (which
is currently 250).
The other change is to make the recommended options for an aggregated
api-server set this limit to 1000. The limit of 250 is annoyingly low
for the use case of many controllers watching objects of Kinds served
by an aggregated api-server reached through the main api-server (in
its mode as a proxy for the aggregated api-server, in which it uses a
single HTTP/2 connection for all calls proxied to that aggregated
api-server).
Fixes#60042
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>.
Update bazelbuild/rules_go, kubernetes/repo-infra, and gazelle dependencies
**What this PR does / why we need it**: updates our bazelbuild/rules_go dependency in order to bump everything to go1.9.4. I'm separating this effort into two separate PRs, since updating rules_go requires a large cleanup, removing an attribute from most build rules.
**Release note**:
```release-note
NONE
```
Jan Safranek
