Commit Graph

148 Commits (b0af0f68b102e37df474c4c21fa46935850f824b)

Author SHA1 Message Date
Kubernetes Submit Queue 87444522d0 Merge pull request #32088 from piosz/fluentd-daemon-set
Automatic merge from submit-queue

Migrated fluentd addon to daemon set

fix #23224
supersedes #23306 

``` release-note
Migrated fluentd addon to daemon set
```
2016-12-15 23:04:40 -08:00
Piotr Szczesniak c00e57789d Added upgrade story from manifest pod to ds 2016-12-15 13:48:32 +01:00
Wojciech Tyczynski 6e336bfab6 Use Get from cache in apiserver in kubelet 2016-12-13 17:14:56 +01:00
Seth Jennings a40b15d8bd error in setNodeStatus func should not abort node status update 2016-12-12 09:29:24 -06:00
Kubernetes Submit Queue 43233caaf0 Merge pull request #37871 from Random-Liu/use-patch-in-kubelet
Automatic merge from submit-queue (batch tested with PRs 36692, 37871)

Use PatchStatus to update node status in kubelet.

Fixes https://github.com/kubernetes/kubernetes/issues/37771.

This PR changes kubelet to update node status with `PatchStatus`.

@caesarxuchao @ymqytw told me that there is a limitation of current `CreateTwoWayMergePatch`, it doesn't support primitive type slice which uses strategic merge.
* I checked the node status, the only primitive type slices in NodeStatus are as follows, they are not using strategic merge:
  * [`ContainerImage.Names`](https://github.com/kubernetes/kubernetes/blob/master/pkg/api/v1/types.go#L2963)
  * [`VolumesInUse`](https://github.com/kubernetes/kubernetes/blob/master/pkg/api/v1/types.go#L2909)
* Volume package is already [using `CreateStrategicMergePath` to generate node status update patch](https://github.com/kubernetes/kubernetes/blob/master/pkg/controller/volume/attachdetach/statusupdater/node_status_updater.go#L111), and till now everything is fine. 

@yujuhong @dchen1107 
/cc @kubernetes/sig-node
2016-12-09 11:29:11 -08:00
Wojciech Tyczynski e8d1cba875 GetOptions in client calls 2016-12-09 09:42:01 +01:00
Kubernetes Submit Queue 362dc81d8e Merge pull request #37713 from yujuhong/log_notready_message
Automatic merge from submit-queue (batch tested with PRs 36736, 35956, 35655, 37713, 38316)

Log the condition when node becomes not ready
2016-12-08 19:51:58 -08:00
Random-Liu beba1ebbf8 Use PatchStatus to update node status in kubelet. 2016-12-08 17:13:59 -08:00
Mike Danese e225625a80 add a configuration for kubelet to register as a node with taints
and deprecate register-schedulable
2016-12-06 10:32:54 -08:00
Clayton Coleman 3454a8d52c
refactor: update bazel, codec, and gofmt 2016-12-03 19:10:53 -05:00
Clayton Coleman 5df8cc39c9
refactor: generated 2016-12-03 19:10:46 -05:00
Wojciech Tyczynski 54d49cb404 While updating NodeStatus, only first get served from cache 2016-12-01 11:07:37 +01:00
Yu-Ju Hong b216080980 Log the condition when node becomes not ready
This improves debuggability.
2016-11-30 12:18:08 -08:00
Pengfei Ni f584ed4398 Fix package aliases to follow golang convention 2016-11-30 15:40:50 +08:00
Chao Xu 5e1adf91df cmd/kubelet 2016-11-23 15:53:09 -08:00
Alexander Block ffce5dbbf4 Fix setNodeAddress in combination with cloud providers
Actually update node.Status.Addresses when the host name was provided by
the cloud provider.
2016-11-07 14:34:34 +01:00
Kubernetes Submit Queue 182a09c3c7 Merge pull request #35526 from justinsb/fix_35521_b
Automatic merge from submit-queue

kubelet bootstrap: start hostNetwork pods before we have PodCIDR

Network readiness was checked in the pod admission phase, but pods that
fail admission are not retried.  Move the check to the pod start phase.

Issue #35409 
Issue #35521
2016-11-06 12:53:14 -08:00
Kubernetes Submit Queue 4b1e36f970 Merge pull request #36190 from dashpole/revert_node_inode_pressure_split
Automatic merge from submit-queue

We only report diskpressure to users, and no longer report inodepressure

See #36180 for more information on why #33218 was reverted.
2016-11-06 03:00:34 -08:00
David Ashpole 9aca40dee6 revert #33218. dont need #36180. We only use diskpressure 2016-11-04 08:29:27 -07:00
Justin Santa Barbara ab6d938247 Don't add duplicate Hostname address
If the cloudprovider returned an address of type Hostname, we shouldn't
add a duplicate one.
2016-11-04 10:00:23 -04:00
Justin Santa Barbara f8eb179c2d Create hostNetwork pods even if network plugin not ready
We do now admit pods (unlike the first attempt), but now we will stop
non-hostnetwork pods from starting if the network is not ready.

Issue #35409
2016-11-04 00:11:55 -04:00
Magnus Kulke b7880e7cd8 Populate NodeHostName status. 2016-11-01 01:09:50 +01:00
Connor Doyle c93646e8da Support opaque integer resource accounting.
- Prevents kubelet from overwriting capacity during sync.
- Handles opaque integer resources in the scheduler.
  - Adds scheduler predicate tests for opaque resources.
- Validates opaque int resources:
  - Ensures supplied opaque int quantities in node capacity,
    node allocatable, pod request and pod limit are integers.
  - Adds tests for new validation logic (node update and pod spec).
- Added e2e tests for opaque integer resources.
2016-10-28 10:15:13 -07:00
Lucas Käldström 1cf00d1ff1 Remove the function of --reconcile-cidr and deprecate it 2016-10-26 20:25:35 +03:00
Jing Xu b02481708a Fix volume states out of sync problem after kubelet restarts
When kubelet restarts, all the information about the volumes will be
gone from actual/desired states. When update node status with mounted
volumes, the volume list might be empty although there are still volumes
are mounted and in turn causing master to detach those volumes since
they are not in the mounted volumes list. This fix is to make sure only
update mounted volumes list after reconciler starts sync states process.
This sync state process will scan the existing volume directories and
reconstruct actual states if they are missing.

This PR also fixes the problem during orphaned pods' directories. In
case of the pod directory is unmounted but has not yet deleted (e.g.,
interrupted with kubelet restarts), clean up routine will delete the
directory so that the pod directoriy could be cleaned up (it is safe to
delete directory since it is no longer mounted)

The third issue this PR fixes is that during reconstruct volume in
actual state, mounter could not be nil since it is required for creating
container.VolumeMap. If it is nil, it might cause nil pointer exception
in kubelet.

Details are in proposal PR #33203
2016-10-25 12:29:12 -07:00
Yu-Ju Hong 94f580ef03 Revert "bootstrap: Start hostNetwork pods even if network plugin not ready" 2016-10-25 08:38:59 -07:00
Kubernetes Submit Queue 3c84164bdf Merge pull request #33347 from justinsb/fix_32900
Automatic merge from submit-queue

bootstrap: Start hostNetwork pods even if network plugin not ready
2016-10-24 01:14:06 -07:00
Wojciech Tyczynski ee73fcdadb Update kubelet_node_status.go 2016-10-22 08:44:25 +02:00
Wojciech Tyczynski ad87989378 Kubelet getting node from apiserver cache before update. 2016-10-21 09:21:39 +02:00
Justin Santa Barbara ad6d842a65 Create hostNetwork pods even if network plugin not ready 2016-10-17 10:12:14 -04:00
Lucas Käldström 348717c50a Remove the flannel experimental overlay 2016-10-04 11:53:53 +03:00
David Ashpole fed3f37eef Split NodeDiskPressure into NodeInodePressure and NodeDiskPressure 2016-10-03 11:42:56 -07:00
Kubernetes Submit Queue 906cb1ce70 Merge pull request #33123 from kokhang/node-ip-cloud-provider
Automatic merge from submit-queue

Node-ip is not used when cloud provider is used

Currently --node-ip in kubelet is not being used when kubelet is configured with a cloud provider. With this fix, kubelet will get a list of IPs from the provider and parse it to return the one that matches node-ip.

This fixes #23568
2016-10-01 02:51:19 -07:00
Steve Leon a9123de9b4 Moving validateNodeIP to kubelet_node_status.go 2016-09-30 14:07:13 -07:00
Justin Santa Barbara 54195d590f Use strongly-typed types.NodeName for a node name
We had another bug where we confused the hostname with the NodeName.

To avoid this happening again, and to make the code more
self-documenting, we use types.NodeName (a typedef alias for string)
whenever we are referring to the Node.Name.

A tedious but mechanical commit therefore, to change all uses of the
node name to use types.NodeName

Also clean up some of the (many) places where the NodeName is referred
to as a hostname (not true on AWS), or an instanceID (not true on GCE),
etc.
2016-09-27 10:47:31 -04:00
Steve Leon 6efa1172f5 Node-ip is not used when cloud provider is used
This fixes #23568
2016-09-20 13:49:16 -07:00
Yu-Ju Hong 7ada99181c Limit the number of names per image reported in the node status 2016-09-16 15:16:08 -07:00
Paul Morie 67387632dc Update node status instead of node in kubelet 2016-09-02 16:24:39 -04:00
Kubernetes Submit Queue 4e1ff53bb2 Merge pull request #31730 from pmorie/kubelet-attach-detach-update
Automatic merge from submit-queue

Make it possible to enable controller-managed attach-detach on existing nodes

Fixes #31673.  Now, if a node already exists with the given name on Kubelet startup, the Kubelet will reconcile the value of the controller-managed-attach-detach annotation so that existing nodes can have this feature turned on and off by changing the Kubelet configuration.

cc @kubernetes/sig-storage @kubernetes/rh-cluster-infra
2016-09-01 07:31:18 -07:00
Paul Morie 1805d30b67 Reconcile value of controller-managed attach-detach annotation on existing nodes in Kubelet startup 2016-08-31 17:04:54 -04:00
Tim St. Clair 3808243b9e
Append "AppArmor enabled" to the Node ready condition message 2016-08-31 09:27:47 -07:00
Paul Morie 3b23b9ba9f Add log message in Kubelet when controller attach/detach is enabled 2016-08-26 12:28:37 -04:00
Paul Morie b91ad76066 Kubelet code move: volume / util 2016-08-22 23:35:11 -04:00
Kubernetes Submit Queue 3dad8f7c06 Merge pull request #29907 from luxas/lookup_ip_better
Automatic merge from submit-queue

[kubelet] Auto-discover node IP if neither cloud provider exists and IP is not explicitly specified

One example where the earlier implementation failed is when running kubelet on CoreOS (bare-metal), where the nameserver is set to `8.8.8.8`. kubelet tries to lookup the node name agains Google DNS, which obviously fails. The kubelet won't recover after that.

The workaround hsa been to set `--hostname-override` to an IP address, but it's quite annoying to try to make a multi-distro way of getting the IP in bash for example. This way is much cleaner.

Refactored the function a little bit at the same time

@vishh @yujuhong @resouer @Random-Liu
2016-08-06 02:26:30 -07:00
Andrey Kurilin 9f1c3a4c56 Fix various typos in kubelet 2016-08-03 01:14:44 +03:00
Lucas Käldström 25d9779f06 Make the lookup function of the node ip address more robust 2016-08-02 14:03:20 +03:00
derekwaynecarr c3324b88a0 Eviction manager observes and acts on disk pressure 2016-07-28 16:01:38 -04:00
Paul Morie 249da77371 Extract kubelet node status into separate file 2016-07-22 01:21:30 -04:00