Commit Graph

36812 Commits

Author SHA1 Message Date
mbohlool
4ed8328e8d Fix a typo in util.sh that resulted in wrong openapi spec generation 2016-10-05 15:59:38 -07:00
Kubernetes Submit Queue
9aac6cfdb6 Merge pull request #34068 from Random-Liu/add-cri-presubmit-test
Automatic merge from submit-queue

CRI: Add presubmit CRI validation test.

For #31459.

This PR adds a new suite for CRI presubmit validation which runs non-flaky, non-serial, non-slow test per-pr.

Except this PR, I'll also change the test-infra side. Ideally, after this is done, we should be be able to trigger CRI validation test per-pr with something like `@k8s-bot cri node e2e test this` and `@k8s-bot cri e2e test this`.

@yujuhong @feiskyer @yifan-gu @freehan 
/cc @kubernetes/sig-node
2016-10-05 15:38:25 -07:00
Kubernetes Submit Queue
05192d9d57 Merge pull request #33319 from juanvallejo/jvallejo/add-option-to-set-nodeport
Automatic merge from submit-queue

Add option to set a service nodeport

**Release note**:
```release-note
Add kubectl --node-port option for specifying the service nodeport
```

This patch adds the option to set a nodeport when creating a NodePort
service. In case of a port allocation error due to a specified port
being out of the valid range, the error now includes the valid
range. If a `--node-port` value is not specified, it defaults to zero, in
which case the allocator will default to its current behavior of
assigning an available port.

This patch also adds a new helper function in `cmd/util/helpers.go` to
retrieve `Int32` cobra flags.

**Example**
```
# create a nodeport service with an invalid port
$ kubectl create service nodeport mynodeport --tcp=8080:7777 --node-port=1
The Service "mynodeport" is invalid: spec.ports[0].nodePort: Invalid
value: 1: provided port is not in the valid range. Valid ports range
from 30000-32767

# create a nodeport service with a valid port
$ kubectl create service nodeport mynodeport --tcp=8080:7777 --node-port=30000
service "mynodeport" created

# create a nodeport service with a port already in use
$ kubectl create service nodeport mynodeport --tcp=8080:7777 --node-port=30000
The Service "mynodeport" is invalid: spec.ports[0].nodePort: Invalid value: 3000: provided port is already allocated

$ kubectl describe service mynodeport
Name:                   mynodeport
Namespace:              default
Labels:                 app=mynodeport
Selector:               app=mynodeport
Type:                   NodePort
IP:                     172.30.81.254
Port:                   8080-7777       8080/TCP
NodePort:               8080-7777       30000/TCP
Endpoints:              <none>
Session Affinity:       None
No events.
```

@fabianofranz
2016-10-05 15:00:32 -07:00
Timothy St. Clair
f2b8b08651 Turn down hootloop logs in priorities. Hit once cluster is near capacity 2016-10-05 16:58:23 -05:00
Kubernetes Submit Queue
bd3664cbef Merge pull request #34000 from wojtek-t/set_cache_size
Automatic merge from submit-queue

Set deserialization cache size based on target memory usage

**Special notes for your reviewer**:

This is the PR we talked about yesterday.

**Release note**:

```release-note
To reduce memory usage to reasonable levels in smaller clusters, kube-apiserver now sets the deserialization cache size based on the target memory usage.
```
2016-10-05 14:23:39 -07:00
Random-Liu
6a4a6e5b6c Add presubmit CRI validation test. 2016-10-05 14:05:05 -07:00
Kubernetes Submit Queue
234c562d4a Merge pull request #33310 from deads2k/controller-04-auto-create-SA
Automatic merge from submit-queue

use service accounts as clients for controllers

Makes it possible for the controller-manager to use service accounts to run individual controllers.  To start, this only enables this feature if this particular controller manager has the power to create service account tokens.  Otherwise, the full-powered client is used instead.  This is a necessary step on the way to subdividing the authority of controllers.

@kubernetes/sig-auth 
@erictune I know you care about this
@ncdc fyi
2016-10-05 13:46:44 -07:00
Jordan Liggitt
3b5c3c286c
Fix wait.JitterUntil 2016-10-05 16:29:06 -04:00
Kubernetes Submit Queue
6f78c0d912 Merge pull request #33951 from pmorie/selinux-overhaul
Automatic merge from submit-queue

Proposal: SELinux enhancements

TLDR: Try to make SELinux support better by not requiring Kubelet directory to be labeled with an SELinux type usable from the container.

cc @kubernetes/sig-node @yifan-gu
2016-10-05 13:09:18 -07:00
Kubernetes Submit Queue
a516ec3085 Merge pull request #34129 from mwielgus/daemonset-client
Automatic merge from submit-queue

Daemonset in federated clientset

Only hack/update-codegen.sh was edited manually.

cc: @nikhiljindal @quinton-hoole @kubernetes/sig-cluster-federation
2016-10-05 13:09:10 -07:00
Kubernetes Submit Queue
87393e1da9 Merge pull request #34106 from kargakis/rollout-status-watch-flag
Automatic merge from submit-queue

kubectl: add watch flag in rollout status

Fixes https://github.com/kubernetes/kubernetes/issues/30627

@kubernetes/kubectl
2016-10-05 13:09:01 -07:00
Kubernetes Submit Queue
debac32d51 Merge pull request #33972 from bowei/flake-issue-27023
Automatic merge from submit-queue

Add a retry loop to KubeletManagerEtcHosts flake

Add a retry loop to KubeletManagerEtcHosts flake
    
    Note: this still makes the test fail if a retry occurs, but
    will give us more information regarding whether or not the
    test flake could be occuring due to delay in mounting of
    /etc/hosts.
2016-10-05 12:27:28 -07:00
deads2k
c30b2efc46 update replicaset lister 2016-10-05 15:20:27 -04:00
Random-Liu
b7b95ccb19 Add serial and benchmark test suite. 2016-10-05 12:03:32 -07:00
Kubernetes Submit Queue
d4088fc47b Merge pull request #34020 from juanvallejo/jvallejo/allow-empty-annotation-values
Automatic merge from submit-queue

Allow empty annotation values

Related downstream issue: https://github.com/openshift/origin/issues/11175

**Release note**:
```release-note
release-note-none
```

Annotations with empty values can be used, for example, in diagnostics
logging. This patch removes the client-side check for empty values in
an annotation key-value pair.

**Before**
```
$ kubectl annotate pod zookeeper-1 node-selector="" --overwrite
error: invalid annotation format: node-selector=
```

**After**
```
$ kubectl annotate pod zookeeper-1 node-selector="" --overwrite
pod "zookeper-1" annotated
```

```
$ kubectl get po/zookeeper-1 --template='{{.metadata.annotations}}'
map[... node-selector: test-label:test]
```

@fabianofranz @liggitt
2016-10-05 11:48:02 -07:00
Kubernetes Submit Queue
da44775a2c Merge pull request #34107 from MrHohn/e2e_sourceip_fix
Automatic merge from submit-queue

Fix #31085, include output checking in retry loop

Fix #31085.

`wget -qO-` might omit the error and `RunHostCmd()` won't be able to catch it sometimes. Verify stdout is not empty within the retry loop to fix. Test passed 1500+ times in a row.

@freehan
2016-10-05 11:02:35 -07:00
Minhan Xia
1a9bafd679 fix bug in DialFromNode where it fail the test when encountered a command execution failure 2016-10-05 11:00:57 -07:00
Marcin
64d5335f98 Daemonset in federated clientset 2016-10-05 19:58:49 +02:00
Minhan Xia
df92825c33 Revert "Revert "move pod networking tests common"" 2016-10-05 10:53:22 -07:00
deads2k
358a57d74a update deployment lister 2016-10-05 13:27:35 -04:00
Bowei Du
d61d3e48ee Add a retry loop to KubeletManagerEtcHosts flake
Note: this still makes the test fail if a retry occurs, but
will give us more information regarding whether or not the
test flake could be occuring due to delay in mounting of
/etc/hosts.
2016-10-05 10:24:21 -07:00
juanvallejo
a872657371
Allow empty annotation values
Annotations with empty values can be used, for example, in diagnostics
logging. This patch removes the client-side check for empty values in
an annotation key-value pair.

**Before**
```
$ kubectl annotate pod zookeeper-1 node-selector="" --overwrite
error: invalid annotation format: node-selector=
```

**After**
```
$ kubectl annotate pod zookeeper-1 node-selector="" --overwrite
pod "zookeper-1" annotated
```

```
$ kubectl get po/zookeeper-1 --template='{{.metadata.annotations}}'
map[... node-selector: test-label:test]
```
2016-10-05 13:17:08 -04:00
deads2k
8ea2acc6a3 use service accounts as clients for controllers 2016-10-05 13:15:16 -04:00
deads2k
c2ed560991 fix watch.until and added listwatch 2016-10-05 13:15:16 -04:00
Kubernetes Submit Queue
617fa91264 Merge pull request #34098 from deads2k/sttts-handlerchain
Automatic merge from submit-queue

Customizable genericapiserver handler chain

Rebase of https://github.com/kubernetes/kubernetes/pull/33662.  This is @sttts, but I rebased it for him since he's out this week.
2016-10-05 09:45:57 -07:00
Dan Williams
e2c96db72b kubenet: make it more apparent that kubenet ignores PodCIDR updates
https://github.com/kubernetes/kubernetes/issues/32900
2016-10-05 11:24:02 -05:00
MrHohn
c61911eedd Fix #31085, include output checking in retry loop 2016-10-05 09:13:19 -07:00
Kubernetes Submit Queue
5caf7199ee Merge pull request #33934 from soltysh/audit_groups
Automatic merge from submit-queue

Add asgroups to audit log

@deads2k || @sttts while reworking origing to use upstream I've noticed we're missing this, ptal
2016-10-05 09:08:23 -07:00
Kubernetes Submit Queue
a6b9e53da5 Merge pull request #34103 from gmarek/long_integration_tests
Automatic merge from submit-queue

Allow setting short value in integration tests

@deads2k
2016-10-05 08:30:01 -07:00
Kubernetes Submit Queue
0798f02391 Merge pull request #33942 from kargakis/sttts-kubectl-run-watch-until
Automatic merge from submit-queue

Switch kubectl to use watch.Until

Squashed version of https://github.com/kubernetes/kubernetes/pull/30235 

@ncdc PTAL

cc: @kubernetes/kubectl
2016-10-05 08:29:54 -07:00
Kubernetes Submit Queue
740458255b Merge pull request #34093 from wojtek-t/fix_index_listing
Automatic merge from submit-queue

Fix call to Index call in listers

This will actually eliminated all logs like this:
```
can not retrieve list of objects using index : object has no meta: object does not implement the Object interfaces
```

And it should also speed up different things (as so far we we basically not using indexes at all.

@kubernetes/sig-api-machinery
2016-10-05 07:45:58 -07:00
gmarek
cb167270bf Allow setting short value in integration tests 2016-10-05 16:36:47 +02:00
Michail Kargakis
7e972f4cbd kubectl: add watch flag in rollout status 2016-10-05 16:33:06 +02:00
Dr. Stefan Schimanski
68cee1d9ac Make genericapiserver handler chain customizable 2016-10-05 10:32:36 -04:00
Dr. Stefan Schimanski
7cfd0150e4 Turn RequestInfoResolver into a RequestInfoFactory 2016-10-05 10:32:36 -04:00
Kubernetes Submit Queue
a7f6557f1d Merge pull request #34074 from asalkeld/annotate-local-option
Automatic merge from submit-queue

Add local option to annotate

**What this PR does / why we need it**:
Add the --local option to be both consistent with other commands and so it can be used with "set selector" and "create service".

**Which issue this PR fixes** 
Related: #7296

**Special notes for your reviewer**:
None

**Release note**:
```release-note
Add a new option "--local" to the `kubectl annotate`
```
2016-10-05 07:09:05 -07:00
Maciej Szulik
9cc2a0454d Add asgroups to audit log 2016-10-05 15:40:11 +02:00
Kubernetes Submit Queue
776af8e7f6 Merge pull request #33990 from asalkeld/local-option-label
Automatic merge from submit-queue

Add local option to the label command

**What this PR does / why we need it**:
Add the --local option to be both consistent with other commands and so it can be used with  "set selector" and "create service".

**Which issue this PR fixes** 
Related: #7296

**Special notes for your reviewer**:
None

**Release note**:
```release-note
Add a new option "--local" to the `kubectl label`
```
2016-10-05 06:32:13 -07:00
Kubernetes Submit Queue
350fafe011 Merge pull request #34023 from mwielgus/fed-deployment
Automatic merge from submit-queue

Add deployment to federation clientset

Only hack/update-codegen.sh was updated manually.

cc: @quinton-hoole @nikhiljindal @kubernetes/sig-cluster-federation
2016-10-05 05:55:21 -07:00
Kubernetes Submit Queue
f932e8d2c1 Merge pull request #34089 from wojtek-t/with_serializable
Automatic merge from submit-queue

Make gets for previous value in watch serializable

Ref #33653
2016-10-05 05:55:14 -07:00
Wojciech Tyczynski
c215bd0eac Fix call to Index call in listers 2016-10-05 14:23:31 +02:00
Kubernetes Submit Queue
7856e46f1b Merge pull request #33967 from xiang90/fix_33772
Automatic merge from submit-queue

scheduler: initialize podsWithAffinity

Without initializing podsWithAffinity, scheduler panics when deleting
a pod from a node that has no pods with affinity ever scheduled to.

Fix #33772
2016-10-05 05:17:51 -07:00
Marcin
34753b6950 Add deployment to federation clientset 2016-10-05 13:43:59 +02:00
Kubernetes Submit Queue
0ad50d2033 Merge pull request #33952 from deads2k/annoyance-specific-integration-test
Automatic merge from submit-queue

specify flags to test-integration

Allows a specific test to be run in test-integration:  ` hack/test-integration.sh auth -test.run=TestKindAuthorization`

@eparis I don't know how good or bad my bash is.
2016-10-05 04:41:01 -07:00
Wojciech Tyczynski
82c360a252 Make gets for previous value in watch serializable 2016-10-05 13:18:31 +02:00
Xiang Li
dc62007478 scheduler: cache.delete deletes the pod from node specified in the cached state 2016-10-05 03:33:53 -07:00
Kubernetes Submit Queue
0331b27f45 Merge pull request #34053 from bprashanth/trim_string
Automatic merge from submit-queue

Check for empty string post trimming

We curl in a retry loop and timeout,  trimming stdout to find endpoint names. When curl hits the timeout, stdout is empty, so we insert the empty string into the received set of endpoints. 

Fixes https://github.com/kubernetes/kubernetes/issues/32684
2016-10-05 02:32:41 -07:00
Kubernetes Submit Queue
33103e576e Merge pull request #33639 from guangxuli/k8s_test_helper
Automatic merge from submit-queue

if test succ remove the tmp file
2016-10-05 01:56:18 -07:00
Kubernetes Submit Queue
cccd3b3234 Merge pull request #33147 from vishh/fix-upgrade-sh
Automatic merge from submit-queue

Fix base image pinning during upgrades via cluster/gce/upgrade.sh

Fixes #33056

cc @timstclair @maisem
2016-10-05 01:19:56 -07:00
Wojciech Tyczynski
2bfcb1a850 Set deserialization cache size based on target memory usage 2016-10-05 10:00:02 +02:00