Commit Graph

19757 Commits

Author SHA1 Message Date
Jordan Liggitt
87e32c7532
Force etcd2 to use application/json, add base64-wrapper decoder as fallback 2017-03-15 11:24:12 -04:00
Kubernetes Submit Queue
fea42bade0 Merge pull request #42854 from vladimirvivien/scaleio-k8s-fix-readOnly
Automatic merge from submit-queue (batch tested with PRs 42854, 43105, 43090)

Update ScaleIO volume plugin default readOnly value

This commit updates the code to set readOnly attribute to be set to false.

**What this PR does / why we need it**:
This PR is a minor fix that updates the default value of `readOnly` attribute to `false`.

**Release note**:

```release-note
NONE
```
2017-03-14 18:44:18 -07:00
Kubernetes Submit Queue
f8dd2569a1 Merge pull request #42991 from jsafrane/fix-default-beta
Automatic merge from submit-queue (batch tested with PRs 42775, 42991, 42968, 43029)

Remove 'beta' from default storage class annotation

I forgot to update default storage class annotation in my storage.k8s.io/v1beta1 -> v1 PRs. Let's fix it before 1.6 is released.

I consider it as a bugfix, in #40088 I already updated the release notes to include non-beta annotation  `storageclass.kubernetes.io/is-default-class`

```release-note
NONE
```


@kubernetes/sig-storage-pr-reviews 
@msau42, please help with merging.
2017-03-14 13:52:41 -07:00
Kubernetes Submit Queue
c11498a34d Merge pull request #42775 from gyliu513/kubectl-taint
Automatic merge from submit-queue (batch tested with PRs 42775, 42991, 42968, 43029)

`kubectl taint` update for `NoExecute`.

**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 #

**Special notes for your reviewer**:
Fixes #42774
**Release note**:

```release-note
```
2017-03-14 13:52:39 -07:00
Kubernetes Submit Queue
a8d8542fc7 Merge pull request #42765 from janetkuo/ds-update-validation-fix
Automatic merge from submit-queue

Add DaemonSet templateGeneration validation and tests, and fix a bunch of validation test errors

For DaemonSet update:
1. Validate that templateGeneration is increased when and only when template is changed
1. Validate that templateGeneration is never decreased
1. Added validation tests for templateGeneration 
1. Fix a bunch of errors in validate tests
   - fake tests: almost all validation test error cases failed on "missing resource version", "name changes", "missing update strategy", "selector/template labels mismatch", not on the real validation we wanted to test
   - some error cases should be success cases

@kargakis @lukaszo @kubernetes/sig-apps-bugs 

*I've verified locally that all DaemonSet e2e tests pass with this change.*
2017-03-14 12:54:49 -07:00
Vladimir Vivien
0715b32439 Update ScaleIO volume plugin default readOnly value
This commit updates the code to set the default value of the readOnly attribute to false.
It also updates the example docs to add full list of supported plugin attributes and doc.
2017-03-14 14:19:48 -04:00
Kubernetes Submit Queue
70b3848bce Merge pull request #42935 from yifan-gu/fix_flock
Automatic merge from submit-queue (batch tested with PRs 42942, 42935)

pkg/util/flock: Fix the flock so it actually locks.

With this PR, the second call to `Acquire()` will block unless the lock is released (process exits).
Also removed the memory mutex in the previous code since we don't need `Release()` here so no need to save and protect the local fd.

Fix #42929.
2017-03-14 10:19:18 -07:00
Kubernetes Submit Queue
6de28fab7d Merge pull request #42942 from vishh/gpu-cont-fix
Automatic merge from submit-queue (batch tested with PRs 42942, 42935)

[Bug] Handle container restarts and avoid using runtime pod cache while allocating GPUs

Fixes #42412

**Background**
Support for multiple GPUs is an experimental feature in v1.6. 
Container restarts were handled incorrectly which resulted in stranding of GPUs
Kubelet is incorrectly using runtime cache to track running pods which can result in race conditions (as it did in other parts of kubelet). This can result in same GPU being assigned to multiple pods.

**What does this PR do**
This PR tracks assignment of GPUs to containers and returns pre-allocated GPUs instead of (incorrectly) allocating new GPUs.
GPU manager is updated to consume a list of active pods derived from apiserver cache instead of runtime cache.
Node e2e has been extended to validate this failure scenario.

**Risk**
Minimal/None since support for GPUs is an experimental feature that is turned off by default. The code is also isolated to GPU manager in kubelet.

**Workarounds**
In the absence of this PR, users can mitigate the original issue by setting `RestartPolicyNever`  in their pods.
There is no workaround for the race condition caused by using the runtime cache though.
Hence it is worth including this fix in v1.6.0.

cc @jianzhangbjz @seelam @kubernetes/sig-node-pr-reviews 

Replaces #42560
2017-03-14 10:19:17 -07:00
Kubernetes Submit Queue
f1e9004da9 Merge pull request #42927 from Random-Liu/fix-kubelet-panic
Automatic merge from submit-queue (batch tested with PRs 42802, 42927, 42669, 42988, 43012)

Fix kubelet panic in cgroup manager.

Fixes https://github.com/kubernetes/kubernetes/issues/42920.
Fixes https://github.com/kubernetes/kubernetes/issues/42875
Fixes #42927 
Fixes #43059

Check the error in walk function, so that we don't use info when there is an error.

@yujuhong @dchen1107 @derekwaynecarr @vishh /cc @kubernetes/sig-node-bugs
2017-03-14 07:31:31 -07:00
Guangya Liu
ab94015ee8 kubectl taint update for NoExecute. 2017-03-14 18:24:40 +08:00
Yifan Gu
a489bd2674 pkg/util/flock: Fix the flock so it actually locks.
With this PR, the second call to `Acquire()` will block unless the lock is released (process exits).
Also removed the memory mutex in the previous code since we don't need `Release()` here so no need to save and protect the local fd.

Fix #42929.
2017-03-13 14:24:59 -07:00
Random-Liu
e6341cc3c7 Fix kubelet panic in cgroup manager. 2017-03-13 12:06:08 -07:00
Vishnu kannan
ad743a922a remove dead code in gpu manager
Signed-off-by: Vishnu kannan <vishnuk@google.com>
2017-03-13 10:58:26 -07:00
Vishnu kannan
ff158090b3 use active pods instead of runtime pods in gpu manager
Signed-off-by: Vishnu kannan <vishnuk@google.com>
2017-03-13 10:58:26 -07:00
Vishnu Kannan
8ed9bff073 handle container restarts for GPUs
Signed-off-by: Vishnu Kannan <vishnuk@google.com>
2017-03-13 10:58:26 -07:00
Jan Safranek
06feaccead Remove 'beta' from default storage class annotation 2017-03-13 12:53:41 +01:00
Kubernetes Submit Queue
e1248bcbbc Merge pull request #42962 from k82cn/fix_min_tolerant_time
Automatic merge from submit-queue

Fixed incorrect result of getMinTolerationTime.

For the following case, `getMinTolerationTime` should return one; but  it returned -1 :
1. for tolerations[0], TolerationSeconds is nil, minTolerationTime is not set 
2. for tolerations[1], it's TolerationSeconds (1) is bigger than `minTolerationTime`, so minTolerationTime is still -1 which means infinite.

```
+		{
+			tolerations: []v1.Toleration{
+				{
+					TolerationSeconds: nil,
+				},
+				{
+					TolerationSeconds: &one,
+				},
+			},
+		},
```
2017-03-12 23:55:39 -07:00
Kubernetes Submit Queue
65ddace3ed Merge pull request #42702 from smarterclayton/printer_owners
Automatic merge from submit-queue

Add pkg/printers OWNERS

Should also include more sig-api-machinery as this will be moving to server side
2017-03-12 21:04:57 -07:00
Klaus Ma
d0e04427d7 Fixed incorrect result of getMinTolerationTime. 2017-03-12 20:21:14 +08:00
Kubernetes Submit Queue
e315c388b2 Merge pull request #42944 from liggitt/patch-defaulting
Automatic merge from submit-queue

Ensure patched objects are defaulted correctly

Restores defaulting behavior for patch API calls removed in e34e1abe33 (diff-517d1b81963bbc7c9b0a16e6eb3c0e2f)

Restores the unit test that ensures we get a defaulted result after applying a patch

Fixes https://github.com/kubernetes/kubernetes/issues/42764
Fixes #42834
2017-03-11 17:49:41 -08:00
Kubernetes Submit Queue
3f660a9779 Merge pull request #42913 from aveshagarwal/master-fix-taint-based-eviction-no-node-cidr
Automatic merge from submit-queue

Fix taint based pod eviction for clusters where controller manager is not running with allocate-node-cidrs set

Fixes https://github.com/kubernetes/kubernetes/issues/42733

In my cluster, I have not set allocate-node-cidr, and It is causing taint based pod eviction to fail. 

@gmarek @kubernetes/sig-scheduling-bugs @davidopp @derekwaynecarr
2017-03-11 14:02:45 -08:00
Kubernetes Submit Queue
8cb14a4f7f Merge pull request #42755 from aveshagarwal/master-fix-default-toleration-seconds
Automatic merge from submit-queue (batch tested with PRs 41794, 42349, 42755, 42901, 42933)

Fix DefaultTolerationSeconds admission plugin

DefaultTolerationSeconds is not working as expected. It is supposed to add default tolerations (for unreachable and notready conditions). but no pod was getting these toleration. And api server was throwing this error:

```
Mar 08 13:43:57 fedora25 hyperkube[32070]: E0308 13:43:57.769212   32070 admission.go:71] expected pod but got Pod
Mar 08 13:43:57 fedora25 hyperkube[32070]: E0308 13:43:57.789055   32070 admission.go:71] expected pod but got Pod
Mar 08 13:44:02 fedora25 hyperkube[32070]: E0308 13:44:02.006784   32070 admission.go:71] expected pod but got Pod
Mar 08 13:45:39 fedora25 hyperkube[32070]: E0308 13:45:39.754669   32070 admission.go:71] expected pod but got Pod
Mar 08 14:48:16 fedora25 hyperkube[32070]: E0308 14:48:16.673181   32070 admission.go:71] expected pod but got Pod
```

The reason for this error is that the input to admission plugins is internal api objects not versioned objects so expecting versioned object is incorrect. Due to this, no pod got desired tolerations and it always showed:

```
Tolerations: <none>
```

After this fix, the correct  tolerations are being assigned to pods as follows:

```
Tolerations:	node.alpha.kubernetes.io/notReady=:Exists:NoExecute for 300s
		node.alpha.kubernetes.io/unreachable=:Exists:NoExecute for 300s
```

@davidopp @kevin-wangzefeng @kubernetes/sig-scheduling-pr-reviews @kubernetes/sig-scheduling-bugs @derekwaynecarr 

Fixes https://github.com/kubernetes/kubernetes/issues/42716
2017-03-10 22:02:18 -08:00
Jordan Liggitt
464db160b4
Ensure patched objects are defaulted correctly 2017-03-10 22:07:10 -05:00
Kubernetes Submit Queue
59aa924a9b Merge pull request #42642 from fraenkel/envfrom
Automatic merge from submit-queue

Invalid environment var names are reported and pod starts

When processing EnvFrom items, all invalid keys are collected and
reported as a single event.

The Pod is allowed to start.

fixes #42583
2017-03-10 17:37:31 -08:00
Kubernetes Submit Queue
9590f694c8 Merge pull request #41830 from irfanurrehman/fed-rbac-1
Automatic merge from submit-queue

[Federation] Kubefed Init should use the right RBAC API version clientset

**What this PR does / why we need it**:
Implements the need as described in https://github.com/kubernetes/kubernetes/issues/41263
**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/41263

**Special notes for your reviewer**:
@madhusudancs @shashidharatd @marun 
cc @kubernetes/sig-federation-bugs

**Release note**:

```
NONE
```
2017-03-10 15:56:47 -08:00
Kubernetes Submit Queue
486ec2b7c9 Merge pull request #42862 from caesarxuchao/sync-warning
Automatic merge from submit-queue (batch tested with PRs 38805, 42362, 42862)

Let GC print specific message for RESTMapping failure

Make the error messages reported in https://github.com/kubernetes/kubernetes/issues/39816 to be more specific, also only print the message once.

I'll also update the garbage collector's doc to clearly state we don't support tpr yet.

We'll wait for the watchable discovery feature (@sttts are you going to work on that?) to land in 1.7, and then enable the garbage collector to handle TPR.

cc @hongchaodeng @MikaelCluseau @djMax
2017-03-10 14:01:23 -08:00
Kubernetes Submit Queue
d2d3884f83 Merge pull request #42362 from soltysh/deployment_generators
Automatic merge from submit-queue (batch tested with PRs 38805, 42362, 42862)

Fix deployment generator after introducing deployments in apps/v1beta1

This PR does two things:

1. Switches all generator to produce versioned objects, to bypass the problem of having an object in multiple versions, which then results in not having stable generator (iow. producing exactly the same object).
2. Introduces new generator for `apps/v1beta1` deployments.

@kargakis @janetkuo ptal

@kubernetes/sig-apps-pr-reviews @kubernetes/sig-cli-pr-reviews ptal

This is a followup to https://github.com/kubernetes/kubernetes/pull/39683, so I'm adding 1.6 milestone.

```release-note
Introduce new generator for apps/v1beta1 deployments
```
2017-03-10 14:01:21 -08:00
Kubernetes Submit Queue
e2218290cf Merge pull request #42444 from jingxu97/Mar/deleteVolume
Automatic merge from submit-queue (batch tested with PRs 42608, 42444)

Return nil when deleting non-exist GCE PD

When gce cloud tries to delete a disk, if the disk could not be found
from the zones, the function should return nil error. This modified behavior is also consistent with AWS
2017-03-10 12:50:24 -08:00
Avesh Agarwal
c3a80719a2 Fix taint based pod eviction for clusters where controller manager
is not running with --allocate-node-cidrs set.
2017-03-10 15:39:21 -05:00
Chao Xu
d7aef0a338 Let GC print specific message for RESTMapping failure 2017-03-10 11:38:57 -08:00
Kubernetes Submit Queue
e261cabb09 Merge pull request #42877 from gmarek/taint_cleanup
Automatic merge from submit-queue (batch tested with PRs 42877, 42853)

Remove unused functions and make logs slightly better

Zero risk cleanup, removing function that are not used anymore, and adding few more logs to help debugging problems.

cc @aveshagarwal
2017-03-10 09:54:21 -08:00
Kubernetes Submit Queue
18ffc95308 Merge pull request #36704 from fabxc/client-metrics2
Automatic merge from submit-queue

Use Prometheus instrumentation conventions

The `System` and `Subsystem` parameters are subject to removal.
(x-ref: https://github.com/prometheus/client_golang/issues/240)

All metrics should use base units, which is seconds in the duration
case.

Counters should always end in `_total` and metrics should avoid
referring to potential label dimensions. Those should rather be
mentioned in the documentation string.

@kubernetes/sig-instrumentation 

Reference docs:
https://prometheus.io/docs/practices/instrumentation/
https://prometheus.io/docs/practices/naming/

**Release note**:
```
Breaking change: Renamed REST client Prometheus metrics to follow the instrumentation conventions ("request_latency_microseconds" -> "rest_client_request_latency_seconds", "request_status_codes" -> "rest_client_requests_total"). Please update your alerting pipeline if you rely on them. 
```
2017-03-10 09:04:18 -08:00
Maciej Szulik
597a359c38 Error out when cronjob generator not specified, but cronjobs are not available 2017-03-10 12:08:01 +01:00
Maciej Szulik
aa4390750c Introduce new generator for apps/v1beta1 deployments 2017-03-10 12:08:01 +01:00
Maciej Szulik
1049dad0a4 Switch generators to use versioned objects 2017-03-10 12:08:01 +01:00
gmarek
fddac63c27 Remove unused functions and make logs slightly better 2017-03-10 11:57:51 +01:00
Kubernetes Submit Queue
c38717b73a Merge pull request #42843 from janetkuo/ds-status-kubectl
Automatic merge from submit-queue

Add new DaemonSetStatus to kubectl printer and describer

@kargakis @lukaszo @kubernetes/sig-apps-pr-reviews @kubernetes/sig-cli-pr-reviews 

```release-note
Add new DaemonSet status fields to kubectl printer and describer. 
```
2017-03-10 01:56:59 -08:00
Fabian Reinartz
2b66b49a2f Use Histogram instead of Summary
A histogram allows to aggregate by labels and calculate more
comprehensive quantiles.
2017-03-10 07:24:38 +01:00
Fabian Reinartz
49e2074f74 Use Prometheus instrumentation conventions
The `System` and `Subsystem` parameters are subject to removal.
(x-ref: https://github.com/prometheus/client_golang/issues/240)

All metrics should use base units, which is seconds in the duration
case.

Counters should always end in `_total` and metrics should avoid
referring to potential label dimensions. Those should rather be
mentioned in the documentation string.
2017-03-10 07:24:38 +01:00
Kubernetes Submit Queue
ab6fecfa3a Merge pull request #42811 from gnufied/validation-no-probe
Automatic merge from submit-queue (batch tested with PRs 42811, 42859)

 Validation PVs for mount options

We are going to move the validation in its own package and we will be calling validation for individual volume types as needed.

Fixes https://github.com/kubernetes/kubernetes/issues/42573
2017-03-09 18:47:52 -08:00
Avesh Agarwal
9f533de80d Fix DefaultTolerationSeconds admission plugin. It was using
versioned object whereas admission plugins operate on internal objects.
2017-03-09 20:24:43 -05:00
Kubernetes Submit Queue
1f5708d460 Merge pull request #42640 from lukaszo/ds-updates-fix
Automatic merge from submit-queue (batch tested with PRs 42024, 42780, 42808, 42640)

kubectl: respect DaemonSet strategy parameters for rollout status

It handles "after-merge" comments from #41116

cc @kargakis @janetkuo 

I will add one more e2e test later. I need to handle some in company stuff.
2017-03-09 16:41:54 -08:00
Kubernetes Submit Queue
7002c53a9c Merge pull request #42808 from ravisantoshgudimetla/nodecontroller_eviction_flake
Automatic merge from submit-queue (batch tested with PRs 42024, 42780, 42808, 42640)

Node controller test flake 39975 with delay for try function

**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #39975 

/cc @ncdc @gmarek @liggitt
2017-03-09 16:41:52 -08:00
Kubernetes Submit Queue
9498a1270f Merge pull request #42024 from luomiao/fix-vsphere-remove-port
Automatic merge from submit-queue

Remove VCenterPort from vsphere cloud provider.

**What this PR does / why we need it**:
Address a bug inside vsphere cloud provider when a port number other than 443 is specified inside the config file.
The url which is used for communicating with govmomi should not include port number.
A port number other than 443 will result in 404 error.
VCenterPort stays in VSphereConfig structure for backward compatibility.

**Which issue this PR fixes** : fixes https://github.com/kubernetes/kubernetes-anywhere/issues/338
2017-03-09 15:59:33 -08:00
Janet Kuo
39857f4865 Add new DaemonSetStatus to kubectl printer and describer 2017-03-09 15:45:17 -08:00
Hemant Kumar
12d6b87894 Validation PVs for mount options
We are going to move the validation in its own package
and we will be calling validation for individual volume types
as needed.
2017-03-09 18:24:37 -05:00
Kubernetes Submit Queue
d790851c8f Merge pull request #42694 from dchen1107/master
Automatic merge from submit-queue (batch tested with PRs 42734, 42745, 42758, 42814, 42694)

Dropped docker 1.9.x support. Changed the minimumDockerAPIVersion to

1.22

cc/ @Random-Liu @yujuhong 

We talked about dropping docker 1.9.x support for a while. I just realized that we haven't really done it yet. 

```release-note
Dropped the support for docker 1.9.x and the belows. 
```
2017-03-09 15:07:00 -08:00
Kubernetes Submit Queue
5a47671614 Merge pull request #42814 from yujuhong/cri-kubemark-3rd-time-is-the-charm
Automatic merge from submit-queue (batch tested with PRs 42734, 42745, 42758, 42814, 42694)

kubemark: enable CRI in the hollow kubelet
2017-03-09 15:06:58 -08:00
Dawn Chen
69eaea2fcc Merge pull request #42779 from dashpole/fix_status
[Bug Fix] Allow Status Updates for Pods that can be deleted
2017-03-09 13:23:00 -08:00
ravisantoshgudimetla
7d444263a5 Change from Micro to Milli for introducing delay 2017-03-09 14:10:28 -05:00