Automatic merge from submit-queue (batch tested with PRs 49992, 48861, 49267, 49356, 49886)
Set default vmodule flag in integration tests
Re-introduce a default glog vmodule flag to the integration test setup.
The default was removed in d08dfb9 because it was hard-coded and
prevented local override. This commit makes the default overridable.
```release-note
NONE
```
/cc @caesarxuchao
Automatic merge from submit-queue (batch tested with PRs 49992, 48861, 49267, 49356, 49886)
remove unused function
**What this PR does / why we need it**:
remove unused function which is not used months ago.
**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
NONE
```
Automatic merge from submit-queue (batch tested with PRs 49992, 48861, 49267, 49356, 49886)
Emit event and retry when fail to start healthz server on kube-proxy
**What this PR does / why we need it**: Enhance kube-proxy's logic when fail to start healthz server.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: From #49263.
**Special notes for your reviewer**:
/assign @thockin @nicksardo @bowei
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 49992, 48861, 49267, 49356, 49886)
Reintegrate aggregation support for OpenAPI
Reintegrating changes of #46734
Changes summary:
- Extracted all OpenAPI specs to new repo `kube-openapi`
- Make OpenAPI spec aggregator to copy and rename any non-requal model (even with documentation change only).
- Load specs when adding APIServices and retry on failure until successful spec retrieval or a 404.
- Assumes all Specs except aggregator's Spec are static
- A re-register of any APIService will result in updating the spec for that service (Suggestion for TPR: they should be registered to aggregator API Server, Open for discussion if any more changes needed for another PR.)
fixes#48548
Automatic merge from submit-queue (batch tested with PRs 49992, 48861, 49267, 49356, 49886)
Correctly handle empty watch event cache
Fixes https://github.com/kubernetes/kubernetes/issues/49956
Introduced by ada60236f7 which did not adjust the oldest available resourceVersion for an empty watch event cache.
Exposed by 74b9ba3b4d, which allowed controllers to get list results from etcd before the watch cache is ready (normally they list with resourceVersion=0 which serves the list request from the watch cache, blocking until it is ready)
When the watch cache had an empty cache of watch events, it currently allows establishing a watch as if it can deliver a watch event for its currently synced resourceVersion. This results in an off-by-one error which can result in a missed watch event.
Scenario:
bob:
1. creates object at resourceVersion=11
sally:
1. does a list API request, gets a list resourceVersion of 10 (just before bob creates the object)
2. starts watch handled by watch cache at resourceVersion=10
Watch cache:
1. initial list gets resourceVersion=11, including the item created by bob
2. when determining the initial watch events to send to sally's watch, there are no watch events in the cache, so no initial watch events are sent.
3. the cache listerwatcher watches etcd starting at resourceVersion=11, so future events are fed into the event cache and to sally's watch
The watch cache should have dropped sally's watch from resourceVersion=10 with a "gone" error, since it can't deliver the watch event for resourceVersion=11. This would force sally to relist (where she would get a list at resourceVersion=11) and rewatch (from resourceVersion=11)
This particularly affects tests that create CRD/TPRs and establish watches on the new types as the storage layer's watch cache is also populating for that type.
```release-note
Fix a bug in watch cache sometimes causing missing events after watch cache initialization.
```
Automatic merge from submit-queue
Add parallelism to GCE cluster upgrade
Fixes https://github.com/kubernetes/kubernetes/issues/48373
Should allow upgrading 500-node cluster (1.6->1.7) in < 1 hr. It currently takes ~1.5 day.
Though it is the duty of the upgrader to choose the right parallelism in order to avoid disrupting too many pods.
/cc @kubernetes/sig-cluster-lifecycle-pr-reviews @kubernetes/sig-scalability-misc @mikedanese @gmarek
Automatic merge from submit-queue (batch tested with PRs 49871, 49422, 49092, 49858, 48999)
ScaleIO Volume Plugin - Volume attribute fixes and updates
**What this PR does / why we need it**:
This is a housekeeping PR for small enhancements and fixes to the ScaleIO volume plugin to address issues:
- Enforcement of fsGroup
- Enable ScaleIO multiple-instance volume mapping
- Tighter validation of PVC parameters
- Injection of default PVC capacity when omitted
- Better alignment of PVC, PV, and volume names for dynamic provisioning
**Special notes for your reviewer**:
**Release note**:
```release-note
Enforcement of fsGroup; enable ScaleIO multiple-instance volume mapping; default PVC capacity; alignment of PVC, PV, and volume names for dynamic provisioning
```
Automatic merge from submit-queue (batch tested with PRs 49871, 49422, 49092, 49858, 48999)
adding kube-apiserver starting option tests
**What this PR does / why we need it**:
There is only one test for `--enable-swagger-ui` option in `cmd/kube-apiserver/app/options/options_test.go`, I have expanded this test with more kube-apiserver starting options.
**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 (batch tested with PRs 49871, 49422, 49092, 49858, 48999)
Refactor logging e2e tests, add new checks
I split existing code into smaller files to simplify the review and future changes
Also, there are new tests for stackdriver logging:
- ingesting system logs from all nodes
- ingesting logs in json format
- ingesting logs in glog format
Automatic merge from submit-queue (batch tested with PRs 49871, 49422, 49092, 49858, 48999)
Add KUBE_COVER help to "make test"
**What this PR does / why we need it**:
Add help info on `KUBE_COVER` to tell contributors how to run test with code coverage.
**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**:
/sig contributor-experience
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue
If error continue for loop
If err does not add continue, type conversion will be error.
If do not add continue, pod. (* V1.Pod) may cause panic to run.
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue
[addon-manager] Remove unneeded annotation codes
**What this PR does / why we need it**:
Clean up addon-manager codes to make it less confusing. The annotation logics is only needed for 1.4->1.5 upgrade.
**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
NONE
```
Automatic merge from submit-queue (batch tested with PRs 49870, 49416, 49872, 49892, 49908)
fix alpha/beta endpoint when api endpoint is specified
fix a bug in alpha/beta compute API endpoint bootstraping when api-endpiont is specified.
```release-note
None
```
Automatic merge from submit-queue (batch tested with PRs 49870, 49416, 49872, 49892, 49908)
Make "kubeadm version" json format output more readable.
**What this PR does / why we need it**:
Add indent to json format output.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
ref: #43750#46598
**Special notes for your reviewer**:
output example:
```
#kubeadm version -o json
{
"clientVersion": {
"major": "1",
"minor": "8+",
"gitVersion": "v1.8.0-alpha.2.1026+d8205661b700c6-dirty",
"gitCommit": "d8205661b700c6f99c33ea0ac1e5ed3e49c202b2",
"gitTreeState": "dirty",
"buildDate": "2017-07-31T12:14:28Z",
"goVersion": "go1.8.3",
"compiler": "gc",
"platform": "linux/amd64"
}
}
```
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 49870, 49416, 49872, 49892, 49908)
FC volume plugin: remove block device at DetachDisk
**What this PR does / why we need it**:
After a volume is unmounted from pod and worker node, and then PV and PVC are deleted, cluster admin or external-provisioner might delete the disk from storage, therefore block device on the node should be cleaned up beforehand.
The photon volume plugin already has same functionality.
**Which issue this PR fixes**: fixes#49392
**Special notes for your reviewer**:
/assign @rootfs
/cc @jsafrane @saad-ali
**Release note**:
```
NONE
```
Automatic merge from submit-queue (batch tested with PRs 49870, 49416, 49872, 49892, 49908)
Renamed zoneNotReadyOrUnreachableTainer to zoneNoExecuteTainer.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: partially fixes#42001
**Release note**:
```release-note
None
```
Automatic merge from submit-queue (batch tested with PRs 49898, 49897, 49919, 48860, 49491)
Fix usage a make(struct, len()) followed by append()
A couple of places in the code we allocate with make() but then use
append(), instead of copy() or direct assignment. This results in a
slice with len() zero elements at the front followed by the expected
data. The correct form for such usage is `make(struct, 0, len())`.
I found these by running:
```
$ git grep -EI -A7 'make\([^,]*, len\(' | grep 'append(' -B7 | grep -v vendor
```
And then manually looking through the results. I'm sure something better
could exist.
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 49898, 49897, 49919, 48860, 49491)
Add basic local volume provisioner e2e tests
**What this PR does / why we need it**:
Adds e2e tests to test local volume provisioner.
**Which issue this PR fixes**: fixes https://github.com/kubernetes/kubernetes/issues/48832
**Special notes for your reviewer**:
- bring up local volume provisioner using bootstrapper
- have provisioner create a volume by creating a directory under discovery path.
- check persistent volume is created
- make a claim on the PV, write some data then delete the claim. Verify volume is cleaned up.
**Release note**:
```release-note
```
@ianchakeres @msau42
Automatic merge from submit-queue (batch tested with PRs 49898, 49897, 49919, 48860, 49491)
gce: make append_or_replace.. atomic
Before this change,
* the final echo is not atomically written to the target file
* two concurrent callers will use the same tempfile
Helps with https://github.com/kubernetes/kubernetes/issues/49895
cc @miekg
Automatic merge from submit-queue (batch tested with PRs 49898, 49897, 49919, 48860, 49491)
gce: extend CLOBBER_CONFIG to support known_tokens.csv
Helps with #49895
Automatic merge from submit-queue
Validate if service has duplicate port
**What this PR does / why we need it**:
Validate if a service has duplicate Spec.Ports.Port (same number with same protocol)
xref #47221
fixes this part:
>It is possible to express a Service with multiple ports blocks with the same number. This is not very useful and may cause trouble for implementations of Services.
**Special notes for your reviewer**:
/cc @thockin @liggitt @mengqiy
@kubernetes/sig-network-pr-reviews
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue
Fixeskubernetes/kubeadm#347: empty node name when joining nodes with kubeadm
Node name discovery failed on `kubeadm join`. If a node name
is not explicitly provided, it will be looked up.
**What this PR does / why we need it**: `kubeadm join` fails because the preflight checks always see an empty node name. This corrects that issue.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixeskubernetes/kubeadm#347
**Special notes for your reviewer**:
**Release note**:
```release-note
kubeadm: Fix join preflight check false negative
```
Automatic merge from submit-queue (batch tested with PRs 46519, 49794, 49720, 49692, 49821)
[route_controller] Emit event when failed to create route
**What this PR does / why we need it**:
When route controller failed to create route for nodes, pod's networking is likely broken. We surface this situation through node condition (https://github.com/kubernetes/kubernetes/blob/v1.7.2/pkg/controller/route/routecontroller.go#L197-L231), but this might not be easily tracked when user have a large cluster with many nodes. Surface the error through event might make things more obvious.
**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 @bowei
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 46519, 49794, 49720, 49692, 49821)
fix missing verb at end of format string
**What this PR does / why we need it**:
A missing verb at end of format string in Errorf call happens in `vendor/k8s.io/client-go/plugin/pkg/client/auth/oidc/oidc_test.go`, This PR fixes this problem.
**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 (batch tested with PRs 46519, 49794, 49720, 49692, 49821)
[make verify] Display list of failed tests to the user at the end
**What this PR does / why we need it**:
Minor improvement to verify all script as it now displays list of failed tests at the end, makes it easier for fixing errors
```
$KUBE_VERIFY_GIT_BRANCH=someBranch hack/make-rules/verify.sh -v -Q
.
.
.
========================
FAILED TESTS
========================
hack/make-rules/../../hack/verify-boilerplate.sh
hack/make-rules/../../hack/verify-godep-licenses.sh
hack/make-rules/../../hack/verify-readonly-packages.sh
```
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes#49845
**Special notes for your reviewer**:
**Release note**:
```release-note
None
```
Automatic merge from submit-queue (batch tested with PRs 46519, 49794, 49720, 49692, 49821)
Log attach detach controller skipping pods at higher priority
This will help us in tracking down problems related to pods
not getting added to desired state of world because of events
arriving out of order or some other problem related to that.
cc @kubernetes/sig-storage-pr-reviews
Automatic merge from submit-queue (batch tested with PRs 46519, 49794, 49720, 49692, 49821)
Federation: Add delaying deliverer to dns controller
**What this PR does / why we need it**:
- if `ensureDNSRecords` returned an error there was no retry, so now introduced an delaying deliverer which would reattempt to do ensureDNSRecords.
- ~~Revamped unit test cases of DNS controller. Added more test cases and increased the test coverage.
This was a leftover job from earlier refactoring PR's.~~
```release-note
NONE
```
/assign @quinton-hoole
cc @marun @madhusudancs @kubernetes/sig-federation-pr-reviews
Automatic merge from submit-queue (batch tested with PRs 49284, 49555, 47639, 49526, 49724)
skip WaitForAttachAndMount for terminated pods in syncPod
Fixes https://github.com/kubernetes/kubernetes/issues/49663
I tried to tread lightly with a small localized change because this needs to be picked to 1.7 and 1.6 as well.
I suspect this has been as issue since we started unmounting volumes on pod termination https://github.com/kubernetes/kubernetes/pull/37228
xref openshift/origin#14383
@derekwaynecarr @eparis @smarterclayton @saad-ali @jwforres
/release-note-none
Automatic merge from submit-queue (batch tested with PRs 49284, 49555, 47639, 49526, 49724)
Supply Portworx StorageClass paramters in volume spec labels for server-side processing
**What this PR does / why we need it**:
This change offloads the requirement of successfully parsing all existing and new portworx volume parameters to it's server-side components. As a result, for fixing bugs in existing volume parameters parsing and adding new support, we will not need to submit a k8s PR.
**Which issue this PR fixes**: fixes#49525
**Release note**:
```release-note
Fix incorrect parsing of io_priority in Portworx volume StorageClass and add support for new paramters.
```
Automatic merge from submit-queue (batch tested with PRs 49284, 49555, 47639, 49526, 49724)
Change pod config to manifest
**What this PR does / why we need it**:
As per https://github.com/kubernetes/kubernetes/pull/46494#discussion_r119675805, change `config` to `manifest` to avoid ambiguous.
**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**:
Since it's a minor fix, there is no issue here.
/cc @mtaufen
**Release note**:
```release-note
NONE
```
Davanum Srinivas