Automatic merge from submit-queue
Add Google cloud KMS service for envelope encryption transformer
This adds the required pieces which will allow addition of KMS based encryption providers (envelope transformer).
For now, we will be implementing it using Google Cloud KMS, but the code should make it easy to add support for any other such provider which can expose Decrypt and Encrypt calls.
Writing tests for Google Cloud KMS Service may cause a significant overhead to the testing framework. It has been tested locally and on GKE though.
Upcoming after this PR:
* Complete implementation of the envelope transformer, which uses LRU cache to maintain decrypted DEKs in memory.
* Track key version to assist in data re-encryption after a KEK rotation.
Development branch containing the changes described above: https://github.com/sakshamsharma/kubernetes/pull/4
Envelope transformer used by this PR was merged in #49350
Concerns #48522
Planned configuration:
```
kind: EncryptionConfig
apiVersion: v1
resources:
- resources:
- secrets
providers:
- kms:
cachesize: 100
configfile: gcp-cloudkms.conf
name: gcp-cloudkms
- identity: {}
```
gcp-cloudkms.conf:
```
[GoogleCloudKMS]
kms-location: global
kms-keyring: google-container-engine
kms-cryptokey: example-key
```
Automatic merge from submit-queue (batch tested with PRs 51298, 51510, 51511)
GCE: Add a fake forwarding rule service
Also add more methods to the address service. These
will be used for testing soon.
Automatic merge from submit-queue (batch tested with PRs 44719, 48454)
check job ActiveDeadlineSeconds
**What this PR does / why we need it**:
enqueue a sync task after ActiveDeadlineSeconds
**Which issue this PR fixes** *:
fixes#32149
**Special notes for your reviewer**:
**Release note**:
```release-note
enqueue a sync task to wake up jobcontroller to check job ActiveDeadlineSeconds in time
```
Automatic merge from submit-queue (batch tested with PRs 44719, 48454)
Fix handling of APIserver errors when saving provisioned PVs.
When API server crashes *after* saving a provisioned PV and before sending
200 OK, the controller tries to save the PV again. In this case, it gets
AlreadyExists error, which should be interpreted as success and not as error.
Especially, a volume that corresponds to the PV should not be deleted in the
underlying storage.
Fixes#44372
```release-note
NONE
```
@kubernetes/sig-storage-pr-reviews
Automatic merge from submit-queue (batch tested with PRs 51425, 51404, 51459, 51504, 51488)
Admit NoNewPrivs for remote and rkt runtimes
**What this PR does / why we need it**:
#51347 is aiming to admit NoNewPrivis for remote container runtime, but it didn't actually solve the problem. See @miaoyq 's comments [here](https://github.com/kubernetes/kubernetes/pull/51347#discussion_r135379446).
This PR always admit NoNewPrivs for runtimes except docker, which should fix the problem.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*:
Fixes#51319.
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 51425, 51404, 51459, 51504, 51488)
Remove previous local storage resource name 'scratch" and "overlay"
Remove previous local storage resource name 'scratch" and "overlay"
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: part of #50818
**Special notes for your reviewer**:
Now local ephemeral storage resource name is "ResourceEphemeralStorage", remove previous names as @vishh suggested in PR #51070
**Release note**:
```release-note
Remove previous local ephemeral storage resource names: "ResourceStorageOverlay" and "ResourceStorageScratch"
```
Automatic merge from submit-queue (batch tested with PRs 50919, 51410, 50099, 51300, 50296)
Remove failure policy from initializer configuration
A few reasons:
* Implementing fail open initializers increases complexity a lot
* We haven't seen strong use cases
* We can always add it back
Automatic merge from submit-queue (batch tested with PRs 50919, 51410, 50099, 51300, 50296)
Remove failure check from deployment controller
@kubernetes/sig-apps-pr-reviews this check is useless w/o automatic rollback so I am removing it.
Automatic merge from submit-queue (batch tested with PRs 50919, 51410, 50099, 51300, 50296)
GCE: Read networkProjectID param
Fixes#48515
/assign bowei
The first commit is the original PR cherrypicked. The master's kubelet isn't provided a cloud config path, so the project is retrieved via instance metadata. In the GKE case, this project cannot be retrieved by the master and caused an error.
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 50919, 51410, 50099, 51300, 50296)
Take mount options to GA by adding PV.spec.mountOptions
**What this PR does / why we need it**: Implements https://github.com/kubernetes/community/pull/771
issue: https://github.com/kubernetes/features/issues/168
**Special notes for your reviewer**:
TODO:
- ~StorageClass mountOptions~
As described in proposal, this adds PV.spec.mountOptions + mountOptions parameter to every plugin that is both provisionable & supports mount options.
(personally, even having done all the work already, i don't agree w/ the proposal that mountOptions should be SC parameter but... :))
**Release note**:
```release-note
Add mount options field to PersistentVolume spec
```
Automatic merge from submit-queue (batch tested with PRs 51471, 50561, 50435, 51473, 51436)
Feature gate initializers field
The metadata.initializers field should be feature gated and disabled by default while in alpha, especially since enforcement of initializer permission that keeps users from submitting objects with their own initializers specified is done via an admission plugin most clusters do not enable yet.
Not gating the field and tests caused tests added in https://github.com/kubernetes/kubernetes/issues/51429 to fail on clusters that don't enable the admission plugin.
This PR:
* adds an `Initializers` feature gate, auto-enables the feature gate if the admission plugin is enabled
* clears the `metadata.initializers` field of objects on create/update if the feature gate is not set
* marks the e2e tests as feature-dependent (will follow up with PR to test-infra to enable the feature and opt in for GCE e2e tests)
```release-note
Use of the alpha initializers feature now requires enabling the `Initializers` feature gate. This feature gate is auto-enabled if the `Initialzers` admission plugin is enabled.
```
Automatic merge from submit-queue (batch tested with PRs 51471, 50561, 50435, 51473, 51436)
Fix inconsistent Prometheus cAdvisor metrics
**What this PR does / why we need it**:
We need this because otherwise kubelet is exposing different sets of Prometheus metrics that randomly include or do not include containers.
See also https://github.com/google/cadvisor/issues/1704; quoting here:
Prometheus requires that all metrics in the same family have the same labels, so we arrange to supply blank strings for missing labels
The function `containerPrometheusLabels()` conditionally adds various metric labels from container labels - pod name, image, etc. However, when it receives the metrics, Prometheus [checks](https://github.com/prometheus/client_golang/blob/master/prometheus/registry.go#L665) that all metrics in the same family have the same label set, and [rejects](https://github.com/prometheus/client_golang/blob/master/prometheus/registry.go#L497) those that do not.
Since containers are collected in (somewhat) random order, depending on which kind is seen first you get one set of metrics or the other.
Changing the container labels function to always add the same set of labels, adding `""` when it doesn't have a real value, eliminates the issue in my testing.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
Fixes#50151
**Special notes for your reviewer**:
I have made the same fix in two places. I am 98% sure the one in `cadvisor_linux.go` isn't used and indeed cannot be used, but have not gone fully down that rabbit-hole.
**Release note**:
```release-note
Fix inconsistent Prometheus cAdvisor metrics
```
Automatic merge from submit-queue (batch tested with PRs 51471, 50561, 50435, 51473, 51436)
Add local storage to downwards API
**Release note**:
```release-note
Add local ephemeral storage to downward API
```
/assign @NickrenREN
Automatic merge from submit-queue (batch tested with PRs 51471, 50561, 50435, 51473, 51436)
Allow zsh completion to be autoloaded by compinit
**What this PR does / why we need it**:
Allows the kubectl zsh autocompletion to be auto loaded by compinit. Had to move the the boilerplate down into the specific shell functions as the compdef needs to be the first line in the definition file.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes#50560
**Special notes for your reviewer**:
**Release note**:
```release-note
kubectl zsh autocompletion will work with compinit
```
Automatic merge from submit-queue (batch tested with PRs 50932, 49610, 51312, 51415, 50705)
Add --request-timeout to kube-apiserver to make global request timeout configurable
**What this PR does / why we need it**:
Make the currently hard coded 60 global request timeout in apiserver configurable via a --request-timeout command line flag.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
https://github.com/kubernetes/kubernetes/issues/51355
**Special notes for your reviewer**:
We plan to backport this to 1.6 and 1.7. To ease the backport work, we've kept this change to a minimum. After this PR we will submit an additional PR to rename `--min-request-timeout` to something that more clearly describes it's purpose, such as `--long-running-request-timeout-min`.
**Release note**:
```release-note
Add --request-timeout to kube-apiserver to make global request timeout configurable.
```
Automatic merge from submit-queue (batch tested with PRs 50932, 49610, 51312, 51415, 50705)
Deprecation warnings for auto detecting cloud providers
**What this PR does / why we need it**:
Adds deprecation warnings for auto detecting cloud providers. As part of the initiative for out-of-tree cloud providers, this feature is conflicting since we're shifting the dependency of kubernetes core into cAdvisor. In the future kubelets should be using `--cloud-provider=external` or no cloud provider at all.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes#50986
**Special notes for your reviewer**:
NOTE: I still have to coordinate with sig-node and kubernetes-dev to get approval for this deprecation, I'm only opening this PR since we're close to code freeze and it's something presentable.
**Release note**:
```release-note
Deprecate auto detecting cloud providers in kubelet. Auto detecting cloud providers go against the initiative for out-of-tree cloud providers as we'll now depend on cAdvisor integrations with cloud providers instead of the core repo. In the near future, `--cloud-provider` for kubelet will either be an empty string or `external`.
```
Automatic merge from submit-queue (batch tested with PRs 50932, 49610, 51312, 51415, 50705)
Add local storage support in Quota
Add local storage(scratch, overlay) support in quota
**Release note**:
```release-note
Add local ephemeral storage support to Quota
```
/cc @ddysher @jingxu97
Automatic merge from submit-queue (batch tested with PRs 50932, 49610, 51312, 51415, 50705)
Implement StatsProvider interface using cadvisor
Ref: https://github.com/kubernetes/kubernetes/issues/46984
- This PR changes the `StatsProvider` interface in `pkg/kubelet/server/stats` so that it can provide container stats from either cadvisor or CRI, and the summary API can consume the stats without knowing how they are provided.
- The `StatsProvider` struct in the newly added package `pkg/kubelet/stats` implements part of the `StatsProvider` interface in `pkg/kubelet/server/stats`.
- In `pkg/kubelet/stats`,
- `stats_provider.go`: implements the node level stats and provides the entry point for this package.
- `cadvisor_stats_provider.go`: implements the container level stats using cadvisor.
- `cri_stats_provider.go`: implements the container level stats using CRI.
- `helper.go`: utility functions shared by the above three components.
- There should be no user visible behaviors change in this PR.
- A follow up PR will implement the StatsProvider interface using CRI.
**Release note**:
```
None
```
/assign @yujuhong
/assign @WIZARD-CXY
When API server crashes *after* saving a provisioned PV and before sending
200 OK, the controller tries to save the PV again. In this case, it gets
AlreadyExists error, which should be interpreted as success and not as error.
Especially, a volume that corresponds to the PV should not be deleted in the
underlying storage.
Prometheus requires that all metrics in the same family have the same
labels, so we arrange to supply blank strings for missing labels
See https://github.com/google/cadvisor/issues/1704
Automatic merge from submit-queue
When faild create pod sandbox record event.
I created pods because of the failure to create a sandbox, but there was no clear message telling me what was the failure, so I wanted to record an event when the sandbox was created.
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue
Always check if default labels on node need to be updated in kubelet
**What this PR does / why we need it**:
Nodes join again but maybe OS/Arch/Instance-Type has changed in the meantime.
In this case the kubelet needs to check if the default labels are still correct and if not it needs to update them.
```release-note
Kubelet updates default labels if those are deprecated
```
Automatic merge from submit-queue (batch tested with PRs 51441, 51356, 51460)
Autodetect kubemark Cloud Provider
**What this PR does / why we need it**:
This is needed for CloudProviderId to be properly
filled for Hollow Nodes.
**Release note**:
```
NONE
```
Automatic merge from submit-queue (batch tested with PRs 51441, 51356, 51460)
Don't update pvc.status.capacity if pvc is already Bound
As discussed here https://github.com/kubernetes/community/pull/657#discussion_r128008128, in order for `pvc.status.Capacity < pv.Spec.Capcity` to be the mechanism for volume filesystem* resize, the pv controller should stop updating pvc.status.Capacity every resync period.
/assign @jsafrane
/sig storage
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 51441, 51356, 51460)
fix the bad position of code comment
**What this PR does / why we need it**:
The position of code comment is wrong and move it to the right position
**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**:
```
NONE
```
Automatic merge from submit-queue
Add volume operation metrics to operation executor and PV controller
This PR implements the proposal for high level volume metrics https://github.com/kubernetes/community/pull/809
**Special notes for your reviewer**:
~Differences from proposal:~ all resolved
~"verify_volume" is now "verify_volumes_are_attached" + "verify_volumes_are_attached_per_node" + "verify_controller_attached_volume." Which of them do we want?~
~There is no "mount_device" metric because the MountVolume operation combines MountDevice and mount (plugin.Setup). Do we want to extract the mount_device metric or is it okay to keep mountvolume as one? For attachable volumes, MountDevice is the actual mount and Setup is a bindmount + setvolumeownership. For unattachable, mountDevice does not occur and Setup is an actual mount + setvolumeownership.~
~PV controller metrics I did not implement following the proposal at all. I did not change goroutinemap nor scheduleOperation. Because provisionClaimOperation does not return an error, so it's impossible for the caller to know if there is actually a failure worth reporting. So I manually create a new metric inside the function according to some conditions.~
@gnufied
I have tested the operationexecutor metrics but not provision & delete. Sample:
**Release note**:
```release-note
Add error count and time-taken metrics for storage operations such as mount and attach, per-volume-plugin.
```
Automatic merge from submit-queue (batch tested with PRs 49861, 50933, 51380, 50688, 51305)
fix v1 validation return error
**What this PR does / why we need it**:
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes#51304
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 49861, 50933, 51380, 50688, 51305)
refactor codes in volume iscsi to improve readability
Signed-off-by: allencloud <allen.sun@daocloud.io>
**What this PR does / why we need it**:
This PR refactors some codes in pkg/volume/iscsi. What is specific, this PR takes advantage of return fast to make codes indent less. As a result the readability of codes will improve a little bit.
What I did:
1. refactor codes in volume iscsi to improve readability.
2. change a keyword of `delete` into `deleteArgs` to reduce ambiguousness.
3. make some variables camel case.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
NONE
**Special notes for your reviewer**:
NONE
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 49861, 50933, 51380, 50688, 51305)
Test loading Kubelet config from a file
**What this PR does / why we need it**:
Adds a test for loading kubelet config from a file, part of improving https://github.com/kubernetes/kubernetes/issues/50217
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 49861, 50933, 51380, 50688, 51305)
Add configurable groups to bootstrap tokens.
**What this PR does / why we need it**:
This change adds support for authenticating bootstrap tokens into a configurable set of extra groups in addition to `system:bootstrappers`. Previously, bootstrap tokens could only ever authenticate to the `system:bootstrappers` group.
Groups are specified as a comma-separated list in the `auth-extra-groups` key of the `bootstrap.kubernetes.io/token` Secret, and must begin with the prefix `system:bootstrapper:` (and match a validation regex that checks against our normal convention). Whether or not any extra groups are configured, `system:bootstrappers` will still be added.
This also adds a `--groups` flag for `kubeadm token create`, which sets the `auth-extra-groups` key on the resulting Secret. The default is to not set the key.
`kubeadm token list` is also updated to include a `EXTRA GROUPS` output column.
**Which issue this PR fixes**: fixes#49306
**Special notes for your reviewer**:
The use case for this is in https://github.com/kubernetes/kubernetes/issues/49306. Comments on the feature itself are probably better over there. It will be part of how HA/self-hosting kubeadm bootstraps new master nodes (post 1.8).
**Release note**:
```release-note
Add support for configurable groups for bootstrap token authentication.
```
cc @luxas @kubernetes/sig-cluster-lifecycle-api-reviews @kubernetes/sig-auth-api-reviews
/kind feature
Automatic merge from submit-queue
simplify disruption controller finder logic
**What this PR does / why we need it**:
Address some comments from https://github.com/kubernetes/kubernetes/pull/45003 and simplify the PDB controller logic as part of issue https://github.com/kubernetes/kubernetes/issues/42284
@enisoc @kargakis @caesarxuchao
Also it feels like we can get rid of the finders all together since with controller ref, each pod has only controller. Let me know if i should remove that finders all together ?
Automatic merge from submit-queue
fix issue(#50937)Fix kubectl get pvc lose volume name
**What this PR does / why we need it**:
closes#50937
**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**:
this should has volumename column.
```
NAME STATUS CAPACITY ACCESS MODES STORAGECLASS AGE
myclaim Bound pv-gcepd2 5Gi ROX slow 35m
myclaim2 Bound pv-gcepd 5Gi ROX slow 25m
```
**Release note**:
```release-note
NONE
```
