Commit Graph

765 Commits

Author SHA1 Message Date
Yu-Ju Hong
8866a894e4 Fix docker api version in kubelet
There are two variables `dockerv110APIVersion` and `dockerV110APIVersion` with
the same purpose, but different values. Remove the incorrect one and fix usage
in the file.
2016-06-07 17:51:36 -07:00
Dr. Stefan Schimanski
12346d5171 Fix seccomp prefix check 2016-06-06 14:08:17 +02:00
Dr. Stefan Schimanski
ad6bd78007 Move /seccomp/ into domain prefix in seccomp annotations
Double slashes are not allowed in annotation keys. Moreover, using the 63
characters of the name component in an annotation key will shorted the space
for the container name.
2016-06-06 14:08:17 +02:00
Random-Liu
49c8683c24 Add timeout for image pulling 2016-06-02 10:49:17 -07:00
k8s-merge-robot
77de942e08 Merge pull request #26451 from Random-Liu/cache_image_history
Automatic merge from submit-queue

Kubelet: Cache image history to eliminate the performance regression

Fix https://github.com/kubernetes/kubernetes/issues/25057.

The image history operation takes almost 50% of cpu usage in kubelet performance test. We should cache image history instead of getting it from runtime everytime.

This PR cached image history in imageStatsProvider and added unit test.

@yujuhong @vishh 
/cc @kubernetes/sig-node 

Mark v1.3 because this is a relatively significant performance regression.

[![Analytics](https://kubernetes-site.appspot.com/UA-36037335-10/GitHub/.github/PULL_REQUEST_TEMPLATE.md?pixel)]()
2016-05-29 20:49:01 -07:00
k8s-merge-robot
344f26ae69 Merge pull request #26145 from Random-Liu/image-pulling-progress
Automatic merge from submit-queue

Kubelet: Periodically reporting image pulling progress in log

Addresses https://github.com/kubernetes/kubernetes/issues/26075#issuecomment-221129896 and https://github.com/kubernetes/kubernetes/pull/26122#issuecomment-221128397.

This PR changes kube_docker_client to log pulling progress every *10* seconds. We can't print all progress messages into the log, because there are too many. So I make it report newest progress every 10 seconds to reduce log spam.
If the image pulling is too slow or stuck, we'll see image pulling progress unchanged or changed little overtime.

The following is the result if I set the reporting interval to *1* second.
```
I0524 00:53:26.189086  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "17b6a9e179d7: Pulling fs layer "
I0524 00:53:27.189082  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "202e40f8bb3a: Download complete "
I0524 00:53:28.189160  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "487bffc61de6: Downloading [=>                                                 ] 1.474 MB/48.35 MB"
I0524 00:53:29.189156  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "487bffc61de6: Downloading [====>                                              ] 3.931 MB/48.35 MB"
I0524 00:53:30.189089  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "487bffc61de6: Downloading [=========>                                         ] 8.847 MB/48.35 MB"
I0524 00:53:31.189089  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "487bffc61de6: Downloading [==================>                                ] 18.19 MB/48.35 MB"
I0524 00:53:32.189076  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "487bffc61de6: Downloading [=======================================>           ] 38.34 MB/48.35 MB"
I0524 00:53:33.189106  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "487bffc61de6: Extracting [=============>                                     ] 12.78 MB/48.35 MB"
I0524 00:53:34.189067  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "487bffc61de6: Extracting [==================================>                ] 33.42 MB/48.35 MB"
I0524 00:53:35.189083  145099 kube_docker_client.go:252] Pulling image "ubuntu:latest": "487bffc61de6: Extracting [==================================================>] 48.35 MB/48.35 MB"
I0524 00:53:35.376667  145099 kube_docker_client.go:254] Finish pulling image "ubuntu:latest": "Status: Downloaded newer image for ubuntu:latest"
```

Ref image pulling related issue #19077.

[![Analytics](https://kubernetes-site.appspot.com/UA-36037335-10/GitHub/.github/PULL_REQUEST_TEMPLATE.md?pixel)]()

@yujuhong @dchen1107 
/cc @kubernetes/sig-node
2016-05-28 13:34:28 -07:00
Random-Liu
52a3d8a19d Add unit test for image history cache 2016-05-27 14:49:48 -07:00
Random-Liu
56bde2df9f Cache image history 2016-05-27 14:49:48 -07:00
Alex Robinson
789b69758e Merge pull request #25688 from sjpotter/rkt_annotations
kubelet: Move common labels out of dockertools package
2016-05-27 11:26:31 -07:00
Alex Mohr
5b1653ec39 Merge pull request #25681 from vishh/lifecycle-probe-logs
Log output of lifecycle hooks on failure
2016-05-26 12:37:02 -07:00
Alex Mohr
4357b8a0a6 Merge pull request #25324 from jfrazelle/add-seccomp
Add Seccomp to Annotations
2016-05-26 10:50:06 -07:00
Shaya Potter
9dd9f2ad65 move common lables used outside of containr runtime out of dockertools
moved labels and their Get functions to types

aliases kubernetes/types import to kubetypes to use kubelet/types as types
2016-05-25 23:10:19 -07:00
k8s-merge-robot
f4122477c2 Merge pull request #25712 from asalkeld/docker-logpath
Automatic merge from submit-queue

Use docker containerInfo.LogPath and not manually constructed path

## Pull Request Guidelines

Since the containerInfo has the LogPath in it, let's use that and
not manually construct the path ourselves. This also makes the code
less prone to breaking if docker change this path.

Fixes #23695
2016-05-25 17:57:59 -07:00
k8s-merge-robot
69b90028cf Merge pull request #26086 from smarterclayton/init_ordering
Automatic merge from submit-queue

Ensure that init containers are preserved during pruning

Pods with multiple init containers were getting the wrong containers
pruned. Fix an error message and add a test.

Fixes #26131
2016-05-25 15:38:39 -07:00
Jess Frazelle
aa8c72adaa seccomp: add annotations and test for docker runtime
Signed-off-by: Jess Frazelle <me@jessfraz.com>
2016-05-24 12:25:21 -07:00
Vishnu kannan
e7721ea8cf Expose errors from libecycle hooks as events for the pod.
Signed-off-by: Vishnu kannan <vishnuk@google.com>
2016-05-24 11:12:53 -07:00
Clayton Coleman
223542690a Ensure that init containers are preserved during pruning
Pods with multiple init containers were getting the wrong containers
pruned. Fix an error message and add a test.
2016-05-24 14:05:55 -04:00
Random-Liu
151d0ab0c1 Periodically reporing image pulling progress in log 2016-05-24 09:58:36 -07:00
Yu-Ju Hong
77dba962db docker: don't set timeout for image pulling requests 2016-05-23 16:17:49 -07:00
Angus Salkeld
2970f2220f Use docker containerInfo.LogPath and not manually constructed path
Since the containerInfo has the LogPath in it, let's use that and
not manually construct the path ourselves. This also makes the code
less prone to breaking if docker change this path.

Fixes #23695
2016-05-23 10:25:37 +10:00
k8s-merge-robot
423a4154be Merge pull request #25062 from dcbw/kubenet-rkt
Automatic merge from submit-queue

Hook rkt kubelet runtime up to network plugins
2016-05-21 00:45:19 -07:00
Vishnu kannan
f884180deb Update kubelet to use per-pod QoS policy.
Signed-off-by: Vishnu kannan <vishnuk@google.com>
2016-05-20 11:52:17 -07:00
k8s-merge-robot
4ea034c284 Merge pull request #25896 from yujuhong/inspect_error
Automatic merge from submit-queue

kubelet: Don't attempt to apply the oom score if container exited already

Containers could terminate before kubelet applies the oom score. This is normal
and the function should not error out.

This addresses #25844 partially.

/cc @smarterclayton @Random-Liu
2016-05-20 06:22:57 -07:00
Yu-Ju Hong
b57fbee557 check pid in oomscoreadj 2016-05-19 12:21:27 -07:00
Yu-Ju Hong
e3e1c7a6a1 kubelet: Don't attempt to apply the oom score if the container is not running
Containers could terminate before kubelet applies the oom score. This is normal
and the function should not error out.
2016-05-19 11:32:36 -07:00
Clayton Coleman
5e4308f91d Update use of Quantity in other classes 2016-05-19 08:41:43 -04:00
Dan Williams
55e6eb2ce3 Allow using netns path instead of container PID to change hairpin mode 2016-05-18 09:25:42 -05:00
Clayton Coleman
205a8b4574 Add init container loading to the kubelet 2016-05-17 00:29:53 -04:00
k8s-merge-robot
59b7b1c550 Merge pull request #24921 from feiskyer/merge-exec
Automatic merge from submit-queue

Remove RunInContainer interface in Kubelet Runtime interface

According to #24689, we should merge RunInContainer and ExecInContainer in the container runtime interface. 

@yujuhong  @kubernetes/sig-node
2016-05-15 02:58:15 -07:00
Yifan Gu
9d5bcf4251 rkt: Add tests for GarbageCollect(). 2016-05-13 15:13:52 -07:00
Jeff Lowdermilk
1fe0e0778f Merge pull request #24080 from tyangliu/fix-infra-network-cfg
Pass dockerOpts by reference to setInfraContainerNetworkConfig
2016-05-13 09:53:39 -07:00
k8s-merge-robot
bb3f5b1768 Merge pull request #23090 from tgraf/ipv6
Automatic merge from submit-queue

Add IPv6 address support for pods - does NOT include services

This allows a container to have an IPv6 address only and extracts the address via nsenter and iproute2 or the docker client directly. An IPv6 address is now correctly reported when describing a pod.

@thockin @kubernetes/sig-network

<!-- Reviewable:start -->
---
This change is [<img src="http://reviewable.k8s.io/review_button.svg" height="35" align="absmiddle" alt="Reviewable"/>](http://reviewable.k8s.io/reviews/kubernetes/kubernetes/23090)
<!-- Reviewable:end -->
2016-05-13 06:32:00 -07:00
Saad Ali
25f37007aa Merge pull request #24846 from pmorie/kubelet-test-loc
Reduce LOC in kubelet tests
2016-05-12 15:52:27 -07:00
k8s-merge-robot
08440b5dcc Merge pull request #24836 from Clarifai/gpu-impl
Automatic merge from submit-queue

WIP v0 NVIDIA GPU support

```release-note
* Alpha support for scheduling pods on machines with NVIDIA GPUs whose kubelets use the `--experimental-nvidia-gpus` flag, using the alpha.kubernetes.io/nvidia-gpu resource 
```

Implements part of #24071 for  #23587

I am not familiar with the scheduler enough to know what to do with the scores. Mostly punting for now.

Missing items from the implementation plan: limitranger, rkt support, kubectl
support and docs

cc @erictune @davidopp @dchen1107 @vishh @Hui-Zhi @gopinatht
2016-05-12 14:04:15 -07:00
Dan Williams
b749902b42 kubenet: set bridge promiscuous mode based on hairpinMode
Closes: https://github.com/kubernetes/kubernetes/issues/23657
Closes: https://github.com/kubernetes/kubernetes/issues/20475
2016-05-12 10:02:33 -05:00
k8s-merge-robot
c4214f743f Merge pull request #24918 from Random-Liu/add-docker-operation-timeout
Automatic merge from submit-queue

Kubelet: Add docker operation timeout

For #23563.
Based on #24748, only the last 2 commits are new.

This PR:
1) Add timeout for all docker operations.
2) Add docker operation timeout metrics
3) Cleanup kubelet stats and add runtime operation error and timeout rate monitoring.
4) Monitor runtime operation error and timeout rate in kubelet perf.

@yujuhong 
/cc @gmarek Because of the metrics change.
/cc @kubernetes/sig-node
2016-05-09 21:51:52 -07:00
Rudi Chiarito
362c763fca WIP v0 NVIDIA GPU support
Implements part of #24071

I am not familiar with the scheduler enough to know what to do with the scores. Punting for now.

Missing items from the implementation plan: limitranger, rkt support, kubectl
support and user docs
2016-05-09 19:20:05 -04:00
k8s-merge-robot
fe135fc251 Merge pull request #24630 from euank/redundant-created
Automatic merge from submit-queue

kubelet: Remove redundant `Container.Created`

As far as I can tell, this has been supplanted by a) the `DockerJSON.CreatedAt` field and b) the
`ContainerStatus.CreatedAt`, where the first is used for creating the
second.

The `.Created` field was only written to as far as I can see.

cc @yifan-gu & @Random-Liu 

Is there any reason we might want to keep this around?
2016-05-08 16:21:05 -07:00
Thomas Liu
019efef8f4 Pass dockerOpts by reference to setInfraContainerNetworkConfig 2016-05-07 14:06:06 -07:00
Andy Goldstein
f091ea5eda Handle image digests in node status and image GC
Start including Docker image digests in the node status and consider image digests during image
garbage collection.
2016-05-07 06:50:51 -04:00
Robert Bailey
a2d8b0af13 Merge pull request #25027 from xiangpengzhao/fix_funcname
Rename a func in manager.go
2016-05-06 20:41:26 -07:00
Random-Liu
148588e6a1 1) Add docker operation timeout metrics.
2) Cleanup kubelet stats and add runtime operation error and timeout
rate monitoring.
3) Monitor runtime operation error and timeout rate in
kubelet perf.
2016-05-06 10:53:13 -07:00
Random-Liu
66678354a0 Add timeout for all docker operation. 2016-05-06 10:53:13 -07:00
derekwaynecarr
7bab6999d4 Allow KillPod to take a gracePeriodOverride 2016-05-06 12:14:43 -04:00
k8s-merge-robot
346ddc52c2 Merge pull request #24748 from Random-Liu/cleanup-with-new-engine-api
Automatic merge from submit-queue

Kubelet: Cleanup with new engine api

Finish step 2 of #23563

This PR:
1) Cleanup go-dockerclient reference in the code.
2) Bump up the engine-api version.
3) Cleanup the code with new engine-api.

Fixes #24076.
Fixes #23809.

/cc @yujuhong
2016-05-06 03:16:53 -07:00
Minhan Xia
04b80f7fb8 rename Status interface to GetPodNetworkStatus 2016-05-04 13:46:31 -07:00
Paul Morie
d1e0e726f2 Reduce LOC in kubelet tests 2016-05-03 22:45:08 -04:00
Xiangpeng Zhao
d27127aaf8 Rename a func in order to be difference with variable
Rename `defaultSecurityOpt` to `getDefaultSecurityOpt`
2016-05-02 03:02:49 +08:00
feisky
831203c19b Remove RunInContainer interface in Kuberlete Runtime interface 2016-05-01 20:21:15 +08:00
Random-Liu
7796b619fd Cleanup the code with new engine-api 2016-04-29 13:34:38 -07:00