Automatic merge from submit-queue
Enable "kick the tires" support for Nvidia GPUs in COS
This PR provides an installation daemonset that will install Nvidia CUDA drivers on Google Container Optimized OS (COS).
User space libraries and debug utilities from the Nvidia driver installation are made available on the host in a special directory on the host -
* `/home/kubernetes/bin/nvidia/lib` for libraries
* `/home/kubernetes/bin/nvidia/bin` for debug utilities
Containers that run CUDA applications on COS are expected to consume the libraries and debug utilities (if necessary) from the host directories using `HostPath` volumes.
Note: This solution requires updating Pod Spec across distros. This is a known issue and will be addressed in the future. Until then CUDA workloads will not be portable.
This PR updates the COS base image version to m59. This is coupled with this PR for the following reasons:
1. Driver installation requires disabling a kernel feature in COS.
2. The kernel API for disabling this interface changed across COS versions
3. If the COS image update is not handled in this PR, then a subsequent COS image update will break GPU integration and will require an update to the installation scripts in this PR.
4. Instead of having to post `3` PRs, one each for adding the basic installer, updating COS to m59, and then updating the installer again, this PR combines all the changes to reduce review overhead and latency, and additional noise that will be created when GPU tests break.
**Try out this PR**
1. Get Quota for GPUs in any region
2. `export `KUBE_GCE_ZONE=<zone-with-gpus>` KUBE_NODE_OS_DISTRIBUTION=gci`
3. `NODE_ACCELERATORS="type=nvidia-tesla-k80,count=1" cluster/kube-up.sh`
4. `kubectl create -f cluster/gce/gci/nvidia-gpus/cos-installer-daemonset.yaml`
5. Run your CUDA app in a pod.
**Another option is to run a e2e manually to try out this PR**
1. Get Quota for GPUs in any region
2. export `KUBE_GCE_ZONE=<zone-with-gpus>` KUBE_NODE_OS_DISTRIBUTION=gci
3. `NODE_ACCELERATORS="type=nvidia-tesla-k80,count=1"`
4. `go run hack/e2e.go -- --up`
5. `hack/ginkgo-e2e.sh --ginkgo.focus="\[Feature:GPU\]"`
The e2e will install the drivers automatically using the daemonset and then run test workloads to validate driver integration.
TODO:
- [x] Update COS image version to m59 release.
- [x] Remove sleep from the install script and add it to the daemonset
- [x] Add an e2e that will run the daemonset and run a sample CUDA app on COS clusters.
- [x] Setup a test project with necessary quota to run GPU tests against HEAD to start with https://github.com/kubernetes/test-infra/pull/2759
- [x] Update node e2e serial configs to install nvidia drivers on COS by default
Packaged the script as a docker container stored in gcr.io/google-containers
A daemonset deployment is included to make it easy to consume the installer
A cluster e2e has been added to test the installation daemonset along with verifying installation
by using a sample CUDA application.
Node e2e for GPUs updated to avoid running on nodes without GPU devices.
Signed-off-by: Vishnu kannan <vishnuk@google.com>
Automatic merge from submit-queue (batch tested with PRs 43726, 43643)
Make a smaller redis image for testing, based on Alpine.
**What this PR does / why we need it**:
This shrinks gcr.io/google_containers/redis from 400MB to 5MB, which should reduce flakes.
**Which issue this PR fixes**:
fixes#43631
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue
test/images/mount-tester: ensure exec binary is o+rx
The `mount-tester` image is currently used as a base layer for other
test images (like `mounttest-user`) which perform uid/gid changes.
However, the binary built in here just follows local umask, and currently is
```
-rwxr-x--- 1 root root 2052704 May 19 2016 mt
```
This commit adds an explicit chmod on the binary to make sure uid/gid
tests can still run it as "others".
The `mount-tester` image is currently used as a base layer for other
test images (like `mounttest-user`) which perform uid/gid changes.
However, the binary built in here just follows local umask, and currently is
```
-rwxr-x--- 1 root root 2052704 May 19 2016 mt
```
This commit adds an explicit chmod on the binary to make sure uid/gid
tests can still run it as "others".
Automatic merge from submit-queue
Move b.gcr.io/k8s_authenticated_test to gcr.io/k8s-authenticated-test
**What this PR does / why we need it**:
Per https://cloud.google.com/container-registry/docs/support/deprecation-notices, b.gcr.io access will be deprecated soon.
I've already mirrored the repo to the location specified in this PR.
Automatic merge from submit-queue
Remove all MAINTAINER statements in the codebase as they are deprecated
**What this PR does / why we need it**:
ref: https://github.com/docker/docker/pull/25466
**Release note**:
```release-note
Remove all MAINTAINER statements in Dockerfiles in the codebase as they are deprecated by docker
```
@ixdy @thockin (who else should be notified?)
Automatic merge from submit-queue (batch tested with PRs 36626, 37294, 37463, 37943, 36541)
Use tmpfs for gluster-server container volumes.
Gluster server needs a filesystem that supports extended attributes for its data. Some distros (Debian, Ubuntu) ship Docker that runs containers on aufs, which does not support extended attributes and therefore Gluster server fails there.
We can use tmpfs for Gluster server data volumes. ~~This expects that host's /tmp is tmpfs, which is true for Debian, Ubuntu, RHEL, CentOS and Fedora.~~
I reworked it to mount tmpfs inside the container, the server pod is privileged.
And *after* this PR is merged and new Gluster server container image is pushed we need to bump the image version in https://github.com/kubernetes/kubernetes/blob/master/test/e2e/volumes.go#L407
Edit: I also fixed Gluster server Dockerfile, it was not working at all.
Gluster server needs a filesystem that supports extended attributes for its
data. Some distros (Debian, Ubuntu) ship Docker that runs containers on aufs,
which does not support extended attributes and therefore Gluster server fails
there.
We can use tmpfs for Gluster server data volumes instead of a directory inside
the container.
Automatic merge from submit-queue
Fixed gcloud command in logs-generator makefile
I grepped through the code looking for `gcloud` and `push` commands and only found one Makefile missing the `--`. I added it.
fixes#33765🐛
Automatic merge from submit-queue
test-images: server address is now configurable
This commit perform changes discussed in #32128
Should be merged before #35301
cc: @pskrzyns
Michael Taufen