Automatic merge from submit-queue
Make big clusters work again after introduction of subnets
This PR does two things:
- make IP aliases automatically pick Node IP Range based on number of Nodes,
- fix logic for starting clusters >4095 Nodes that was broken by introduction of subnets,
cc @wojtek-t @shyamjvs
```release-note
Setting env var ENABLE_BIG_CLUSTER_SUBNETS=true will allow kube-up.sh to start clusters bigger that 4095 Nodes on GCE.
```
Ref https://github.com/kubernetes/kubernetes/issues/47344
ip-masq-agent when issuing an iptables-save will read
any configured iptables on the node. This means that
the ip-masq-agent's memory requirements would grow
with the number of iptables (i.e. services) on the node.
Disable ip-masq-agent in GCE
setting the non-masq-cidr in the kubelet to 0.0.0.0/0 we
need to ensure the ip-masq-agent runs.
Add node label pre-req back to ip-masq-agent.
Make gce test consistent with gce default scripts.
Automatic merge from submit-queue (batch tested with PRs 47530, 47679)
Use cos-stable-59-9460-64-0 instead of cos-beta-59-9460-20-0.
Remove dead code that has now moved to another repo as part of #47467
**Release note**:
```release-note
NONE
```
/sig node
- It contains a fix for ipaliasing.
- It contains a fix which decouples GPU driver installation from kernel
version.
Remove dead code that has now moved to another repo as part of #47467
This will create the necessary certificates.
On GCE is will upload those certificates to Metadata.
They are then pulled down on to the kube-apiserver.
They are written to the /etc/src/kubernetes/pki directory.
Finally they are loaded vi the appropriate command line flags.
The requestheader-client-ca-file can be seen by running the following:-
kubectl get ConfigMap extension-apiserver-authentication
--namespace=kube-system -o yaml
Minor bug fixes.
Made sure AGGR_MASTER_NAME is set up in all configs.
Clean up variable names.
Added additional requestheader configuration parameters.
Added check so that if there is no Aggregator CA contents we won't start
the aggregator with the relevant flags.
Working on fixing https://github.com/kubernetes/kubernetes/issues/43716.
This will create the necessary certificates.
On GCE is will upload those certificates to Metadata.
They are then pulled down on to the kube-apiserver.
They are written to the /etc/src/kubernetes/pki directory.
Finally they are loaded vi the appropriate command line flags.
The requestheader-client-ca-file can be seen by running the following:-
kubectl get ConfigMap extension-apiserver-authentication
--namespace=kube-system -o yaml
Minor bug fixes.
Made sure AGGR_MASTER_NAME is set up in all configs.
Clean up variable names.
Added additional requestheader configuration parameters.
Automatic merge from submit-queue
Adding a metadata proxy addon
**What this PR does / why we need it**: adds a metadata server proxy daemonset to hide kubelet secrets.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: this partially addresses #8867
**Special notes for your reviewer**:
**Release note**: the gce metadata server can be hidden behind a proxy, hiding the kubelet's token.
```release-note
The gce metadata server can be hidden behind a proxy, hiding the kubelet's token.
```
Automatic merge from submit-queue
Add initializer support to admission and uninitialized filtering to rest storage
Initializers are the opposite of finalizers - they allow API clients to react to object creation and populate fields prior to other clients seeing them.
High level description:
1. Add `metadata.initializers` field to all objects
2. By default, filter objects with > 0 initializers from LIST and WATCH to preserve legacy client behavior (known as partially-initialized objects)
3. Add an admission controller that populates .initializer values per type, and denies mutation of initializers except by certain privilege levels (you must have the `initialize` verb on a resource)
4. Allow partially-initialized objects to be viewed via LIST and WATCH for initializer types
5. When creating objects, the object is "held" by the server until the initializers list is empty
6. Allow some creators to bypass initialization (set initializers to `[]`), or to have the result returned immediately when the object is created.
The code here should be backwards compatible for all clients because they do not see partially initialized objects unless they GET the resource directly. The watch cache makes checking for partially initialized objects cheap. Some reflectors may need to change to ask for partially-initialized objects.
```release-note
Kubernetes resources, when the `Initializers` admission controller is enabled, can be initialized (defaulting or other additive functions) by other agents in the system prior to those resources being visible to other clients. An initialized resource is not visible to clients unless they request (for get, list, or watch) to see uninitialized resources with the `?includeUninitialized=true` query parameter. Once the initializers have completed the resource is then visible. Clients must have the the ability to perform the `initialize` action on a resource in order to modify it prior to initialization being completed.
```
Automatic merge from submit-queue (batch tested with PRs 46239, 46627, 46346, 46388, 46524)
Configure NPD version through env variable
This lets user specify NPD version to be installed with kubernetes.
Automatic merge from submit-queue (batch tested with PRs 46299, 46309, 46311, 46303, 46150)
Create a subnet for reserving the service cluster IP range
This will be done if IP aliases is enabled on GCP.
```release-note
NONE
```
Automatic merge from submit-queue
Update Calico add-on
**What this PR does / why we need it:**
Updates Calico to the latest version using self-hosted install as a DaemonSet, removes Calico's dependency on etcd.
- [x] Remove [last bits of Calico salt](175fe62720/cluster/saltbase/salt/calico/master.sls (L3))
- [x] Failing on the master since no kube-proxy to access API.
- [x] Fix outgoing NAT
- [x] Tweak to work on both debian / GCI (not just GCI)
- [x] Add the portmap plugin for host port support
Maybe:
- [ ] Add integration test
**Which issue this PR fixes:**
https://github.com/kubernetes/kubernetes/issues/32625
**Try it out**
Clone the PR, then:
```
make quick-release
export NETWORK_POLICY_PROVIDER=calico
export NODE_OS_DISTRIBUTION=gci
export MASTER_SIZE=n1-standard-4
./cluster/kube-up.sh
```
**Release note:**
```release-note
The Calico version included in kube-up for GCE has been updated to v2.2.
```
KUBE_GCE_ENABLE_IP_ALIASES=true will enable allocation of PodCIDR ips
using the ip alias mechanism rather than using routes.
NODE_IP_RANGE will control the node instance IP cidr
KUBE_GCE_IP_ALIAS_SIZE controls the size of each podCIDR
IP_ALIAS_SUBNETWORK controls the name of the subnet created for the cluster
export functions from pkg/api/validation
add settings API
add settings to pkg/registry
add settings api to pkg/master/master.go
add admission control plugin for pod preset
add new admission control plugin to kube-apiserver
add settings to import_known_versions.go
add settings to codegen
add validation tests
add settings to client generation
add protobufs generation for settings api
update linted packages
add settings to testapi
add settings install to clientset
add start of e2e
add pod preset plugin to config-test.sh
Signed-off-by: Jess Frazelle <acidburn@google.com>
Automatic merge from submit-queue (batch tested with PRs 40932, 41896, 41815, 41309, 41628)
enable DefaultTolerationSeconds admission controller by default
**What this PR does / why we need it**:
Continuation of PR #41414, enable DefaultTolerationSeconds admission controller by default.
**Which issue this PR fixes**:
fixes: #41860
related Issue: #1574, #25320
related PRs: #34825, #41133, #41414
**Special notes for your reviewer**:
**Release note**:
```release-note
enable DefaultTolerationSeconds admission controller by default
```
Automatic merge from submit-queue (batch tested with PRs 40971, 41027, 40709, 40903, 39369)
Bump GCI to gci-beta-56-9000-80-0
cc/ @Random-Liu @adityakali
Changelogs since gci-dev-56-8977-0-0 (currently used in Kubernetes):
- "net.ipv4.conf.eth0.forwarding" and "net.ipv4.ip_forward" may get reset to 0
- Track CVE-2016-9962 in Docker in GCI
- Linux kernel CVE-2016-7097
- Linux kernel CVE-2015-8964
- Linux kernel CVE-2016-6828
- Linux kernel CVE-2016-7917
- Linux kernel CVE-2016-7042
- Linux kernel CVE-2016-9793
- Linux kernel CVE-2016-7039 and CVE-2016-8666
- Linux kernel CVE-2016-8655
- Toolbox: allow docker image to be loaded from local tarball
- Update compute-image-package in GCI
- Change the product name on /etc/os-release (to COS)
- Remove 'dogfood' from HWID_OVERRIDE in /etc/lsb-release
- Include Google NVME extensions to optimize LocalSSD performance.
- /proc/<pid>/io missing on GCI (enables process stats accounting)
- Enable BLK_DEV_THROTTLING
cc/ @roberthbailey @fabioy for GKE cluster update
This is to facilitate GCI tip vs. K8s tip testing; we need to
dynamically set the version of GCI to stay current with their
latest canary (latest of the "gci-base" prefixed images).
There is a concern that some GCE users may be running automation that
(a) turns up ephemeral clusters and (b) always uses the latest K8s
release. If any of these workloads fall outside the set supported on
GCI, cutting the release will break the automation. We are therefore
delaying this change until we have provided sufficient warning.
Automatic merge from submit-queue
Add gcl cluster logging test
This PR changes default logging destination for tests to gcp and adds test for cluster logging using google cloud logging
Fix#20760
Automatic merge from submit-queue
Update GCI_VERSION to gci-dev-55-8866-0-0
Update GCI base image:
Change log:
* Built-in kubernetes updated to v1.4.0
* Enabled VXLAN and IP_SET config options in kernel to support some networking tools
* OpenSSL CVE fixes
```release-note
Update GCI base image:
* Enabled VXLAN and IP_SET config options in kernel to support some networking tools (ebtools)
* OpenSSL CVE fixes
```
cc/ @kubernetes/goog-image cc/ @dchen1107
Automatic merge from submit-queue
Nodefs becomes imagefs on GCI
Kubelet cannot identify rootfs correctly
For #33444
```release-note
Enforce Disk based pod eviction with GCI base image in Kubelet
```
Signed-off-by: Vishnu kannan <vishnuk@google.com>
Changelog:
* Built-in kubernetes updated to v1.4.0
* Enabled VXLAN and IP_SET config options in kernel to support some networking tools
* OpenSSL CVE fixes
Automatic merge from submit-queue
Update the containervm image to the latest one (container-v1-3-v20160…
Node e2e is running with old containervm image which only has docker 1.9.1. This pr fixed such issue.
Automatic merge from submit-queue
Bump up GCI version.
```release-note
Upgrading Container-VM base image for k8s on GCE. Brief changelog as follows:
- Fixed performance regression in veth device driver
- Docker and related binaries are statically linked
- Fixed the issue of systemd being oom-killable
```
Fixes#32596
This needs a cherrypick into v1.4 release branch because it is fixing v1.4 release blocking issues. This patch is easy and safe to rollback in case of emergencies.
@vishh can you please review?
Fixes#32596 and many other issues.
cc/ @kubernetes/goog-image FYI
Brief changelog compared to gci-dev-54-8743-3-0:
- Fixed performance regression in veth device driver
- Docker and related binaries are statically linked
- Fixed the issue of systemd being oom-killable
- Updated built-in kubelet version to 1.3.7
- add ethtool and ebtables binaries expected by kubelet
Fixes#32596
Automatic merge from submit-queue
Implemented KUBE_DELETE_NODES flag in kube-down.
Implemented KUBE_DELETE_NODES flag in kube-down script.
It prevents removal of nodes when shutting down a HA master replica.
Automatic merge from submit-queue
Enable kubelet eviction whenever inodes free is < 5% on GCE
This is a pre-req for enabling inodes based evictions in GKE.
Automatic merge from submit-queue
rkt: Update kube-up rkt version to v1.14.0
cc @kubernetes/sig-rktnetes
This should have been included in #31286 (whoops).
This is a bugfix that I propose for v1.4 inclusion.
Automatic merge from submit-queue
Enable Rescheduler by default
Rescheduler is stable - e2e test is passing constantly for >1week.
ref #29023
```release-note
Rescheduler which ensures that critical pods are always scheduled enabled by default in GCE.
```
Automatic merge from submit-queue
Pick a specific GCI version by default on GCE.
Prior to this change, a K8s branch (master as well as release) was
pinned to a GCI milestone. It would pick up the latest GCI release on
that milestone at the time of cluster creation. The rationale was the
K8s users would automatically get the bug fixes in newer versions of
GCI. However in practice, it makes the runtime environment
non-deterministic, and lack of continuous e2e tests mean we would run
into breakages sooner or later.
With this change, each K8s release will pick a specific version
of GCI by default (similar to how the Debian-based container-vm gets used).
Users can override the default version through KUBE_GCE_MASTER_IMAGE and
KUBE_GCE_NODE_IMAGE environment variables.
We expect the default GCI version will be updated relatively frequently stay
updated with newer GCI releases. We can also automate the process to
automatically bump the hard-coded GCI version in future.
@vishh @adityakali can you please review?
cc @kubernetes/goog-image FYI
Prior to this change, a K8s branch (master as well as release) was
pinned to a GCI milestone. It would pick up the latest GCI release on
that milestone at the time of cluster creation. The rationale was the
K8s users would automatically get the bug fixes in newer versions of
GCI. However in practice, it makes the runtime environment
non-deterministic, and lack of continuous e2e tests mean we would run
into breakages sooner or later.
With this change, each K8s release will pick a specific version
of GCI by default (similar to how the Debian-based container-vm gets used).
Users can override the default version through KUBE_GCE_MASTER_IMAGE and
KUBE_GCE_NODE_IMAGE environment variables.
We expect the default GCI version will be updated relatively frequently stay
updated with newer GCI releases. We can also automate the process to
automatically bump the hard-coded GCI version in future.
Automatic merge from submit-queue
Add admission controller for default storage class.
The admission controller adds a default class to PVCs that do not require any
specific class. This way, users (=PVC authors) do not need to care about
storage classes, administrator can configure a default one and all these PVCs
that do not care about class will get the default one.
The marker of default class is annotation "volume.beta.kubernetes.io/storage-class", which must be set to "true" to work. All other values (or missing annotation) makes the class non-default.
Based on @thockin's code, added tests and made it not to reject a PVC when no class is marked as default.
.
@kubernetes/sig-storage
The admission controller adds a default class to PVCs that do not require any
specific class. This way, users (=PVC authors) do not need to care about
storage classes, administrator can configure a default one and all these PVCs
that do not care about class will get the default one.
* Add a pillar for hostname (because even if there's a good Salt
function for it, I don't trust it to return the short hostname)
* Move INITIAL_ETCD_CLUSTER to just the GCE turn-up
* Remove the master_name, which isn't needed as a pillar
Automatic merge from submit-queue
Add Calico as policy provider in GCE
Adds Calico as policy provider to GCE, enforcing the extensions/v1beta1 NetworkPolicy API.
Still to do:
- [x] Enable NetworkPolicy API when POLICY_PROVIDER is provided.
- [x] Fix CNI plugin, policy controller versions.
CC @thockin - does this general approach look good?
Automatic merge from submit-queue
Prep for continuous Docker validation test
```release-note
Add a test config variable to specify desired Docker version to run on GCI.
```
We want to continuously validate Docker releases (#25215), on GCI. This change
adds a new test config variable, `KUBE_GCI_DOCKER_VERSION`, through which we can
specify which version of Docker we want to run on the master and nodes. This
change also patches the Jenkins e2e-runner with the ability to fetch the latest
Docker (pre)release, and sets the aforementioned variable accordingly.
Tested on my local Jenkins instance that was able to start a cluster with the latest Docker version (different from installed version) running on both master and nodes.
@dchen1107 Can you review?
cc/ @andyzheng0831 for changes in `cluster/gce/gci/helper.sh`, and @ixdy @spxtr for changes to the Jenkins e2e-runner
cc/ @kubernetes/goog-image
Automatic merge from submit-queue
Re-enable node problem detector by default
Re-enable node problem detector started in gce cluster by default.
For now, in the master node, the node problem detector will be started and do nothing (see https://github.com/kubernetes/node-problem-detector/pull/13).
But in fact, in my test cluster, the master has no extra cpu to run the node problem detector, so node problem detector is started on all nodes except master, which is what we want but not expected...
@dchen1107
/cc @kubernetes/sig-node
/cc @andyzheng0831 for the gci script change.
[]()
We want to continuously validate Docker releases (#25215), on GCI. This change
adds a new test config variable, `KUBE_GCI_DOCKER_VERSION`, through which we can
specify which version of Docker we want to run on the master and nodes. This
change also patches the Jenkins e2e-runner with the ability to fetch the latest
Docker (pre)release, and sets the aforementioned variable accordingly.
This change recovers some of the side effects of
https://github.com/kubernetes/kubernetes/pull/26197, i.e., keeps the defaults of
`NODE_IMAGE` and `NODE_IMAGE_PROJECT` to `MASTER_IMAGE` and
`MASTER_IMAGE_PROJECT`, for backward compatibility. Although it keeps
`OS_DISTRIBUTION` defaulting to `gci`, the default settings of these vars are
moved to `cluster/gce/util.sh` and conditioned on `OS_DISTRIBUTION==gci`.
Automatic merge from submit-queue
Prepull images in e2e
Quick and dirty image puller because the SQ stalled multiple times just *today* on image pull flake (https://github.com/kubernetes/kubernetes/issues/25277).
@kubernetes/sig-node @kubernetes/sig-testing wdyt?
Automatic merge from submit-queue
Add node problem detector as an addon pod.
```release-note
Introduce a new add-on pod NodeProblemDetector.
NodeProblemDetector is a DaemonSet running on each node, monitoring node health and reporting
node problems as NodeCondition and Event. Currently it already supports kernel log monitoring, and
will support more problem detection in the future. It is enabled by default on gce now.
```
This PR enables NodeProblemDetector as an add-on pod.
/cc @mikedanese @kubernetes/sig-node
[]()
Allow the gcr.io/google_containers registry to be overridden
regionally by just blasting a new KUBE_ADDON_REGISTRY out. Instead of
adding every addon to Salt and asking all of the other consumers
(Trusty, Juju, Mesos, etc) to change, just script the sed ourselves.
This is probably the 9th grossest thing I've ever done, but it works
well, and it works quickly. I kind of wish it didn't.
* In kube-up.sh, create a staging bucket with a location nearest the
zone being created. If new variable RELEASE_REGION_FALLBACK is set
(default false), create multiple buckets and stage to fallback
URLs. (In open source, this path is primarily for testing.)
* In configure-vm.sh, split the URL env variables by comma (if any
extra are present) and retry on the fallback URLs. Also factor the
hash checking into this path rather than outside, since a corrupt
release in a particular geo can be retried in a different geo.
* Remove the local already-staged .tar.gz checks. They've caused
several issues along the way, and with this code path become virtually
unmaintainable. (I could add a sentinel for each bucket it's possibly
staged to, but ew.)
David Ashpole