Automatic merge from submit-queue
Fix nil pointer issue when making mounts for container
When rebooting one of the nodes in my colleague's cluster, two panics were discovered:
```
E1216 04:07:00.193058 2394 runtime.go:52] Recovered from panic: "invalid memory address or nil pointer dereference" (runtime error: invalid memory address or nil pointer dereference)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/util/runtime/runtime.go:58
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/util/runtime/runtime.go:51
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/util/runtime/runtime.go:41
/usr/local/go/src/runtime/asm_amd64.s:472
/usr/local/go/src/runtime/panic.go:443
/usr/local/go/src/runtime/panic.go:62
/usr/local/go/src/runtime/sigpanic_unix.go:24
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/kubelet.go:1313
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/kubelet.go:1473
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/dockertools/docker_manager.go:1495
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/dockertools/docker_manager.go:2125
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/dockertools/docker_manager.go:2093
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/kubelet.go:1971
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/kubelet.go:530
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/pod_workers.go:171
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/pod_workers.go:154
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/pod_workers.go:215
/usr/local/go/src/runtime/asm_amd64.s:1998
E1216 04:07:00.275030 2394 runtime.go:52] Recovered from panic: "invalid memory address or nil pointer dereference" (runtime error: invalid memory address or nil pointer dereference)
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/util/runtime/runtime.go:58
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/util/runtime/runtime.go:51
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/util/runtime/runtime.go:41
/usr/local/go/src/runtime/asm_amd64.s:472
/usr/local/go/src/runtime/panic.go:443
/usr/local/go/src/runtime/panic.go:62
/usr/local/go/src/runtime/sigpanic_unix.go:24
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/server/stats/volume_stat_caculator.go:98
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/kubelet/server/stats/volume_stat_caculator.go:63
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/util/wait/wait.go:86
/go/src/k8s.io/kubernetes/_output/dockerized/go/src/k8s.io/kubernetes/pkg/util/wait/wait.go:87
/usr/local/go/src/runtime/asm_amd64.s:1998
```
kubectl version
```
Client Version: version.Info{Major:"1", Minor:"3", GitVersion:"v1.3.8", GitCommit:"693ef591120267007be359f97191a6253e0e4fb5", GitTreeState:"clean", BuildDate:"2016-09-28T03:03:21Z", GoVersion:"go1.6.2", Compiler:"gc", Platform:"linux/amd64"}
Server Version: version.Info{Major:"1", Minor:"3", GitVersion:"v1.3.8", GitCommit:"693ef591120267007be359f97191a6253e0e4fb5", GitTreeState:"clean", BuildDate:"2016-09-28T02:52:25Z", GoVersion:"go1.6.2", Compiler:"gc", Platform:"linux/amd64"}
```
The second panic had already been fixed by #33616 and #34251. Not sure what caused the first nil pointer issue and whether it has been fixed yet in the master branch. Just fix it by ignoring the nil pointer when making mounts.
cc @jingxu97 @yujuhong
Automatic merge from submit-queue (batch tested with PRs 45110, 45148)
write HostAliases to hosts file
**What this PR does / why we need it**: using the PodSpec's `HostAliases`, we write entries into the Kubernetes-managed hosts file.
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes#43632
**Special notes for your reviewer**:
Previous PRs in this series:
- https://github.com/kubernetes/kubernetes/pull/44572 isolates the logic of creating the file and writing the file
- https://github.com/kubernetes/kubernetes/pull/44641 introduces the `HostAliases` field in PodSpec along with validations
**Release note**:
```release-note
PodSpec's `HostAliases` now write entries into the Kubernetes-managed hosts file.
```
@thockin @yujuhong
Thanks for reviewing!
Automatic merge from submit-queue (batch tested with PRs 45110, 45148)
Make timeouts in the Kubelet slightly offset to aid debugging
Several of these loops overlap, and when they are the reason a failure
is happening it is difficult to sort them out. Slighly misalign these
loops to make their impact obvious.
We are seeing exactly 2 minute pod worker timeouts in a wide range of test flake scenarios, and I want to be confident we know exactly which one is the culprit.
Automatic merge from submit-queue (batch tested with PRs 41583, 45117, 45123)
Implement shared PID namespace in the dockershim
**What this PR does / why we need it**: Defaults the Docker CRI to using a shared PID namespace for pods. Implements proposal in https://github.com/kubernetes/community/pull/207 tracked by #1615.
//cc @dchen1107 @vishh @timstclair
**Special notes for your reviewer**: none
**Release note**:
```release-note
Some container runtimes share a process (PID) namespace for all containers in a pod. This will become the default for Docker in a future release of Kubernetes. You can preview this functionality if running with the CRI and Docker 1.13.1 by enabling the --experimental-docker-enable-shared-pid kubelet flag.
```
Automatic merge from submit-queue (batch tested with PRs 45033, 44961, 45021, 45097, 44938)
Cleanup orphan logging that goes on in the sync loop.
**What this PR does / why we need it**:
Fixes#44937
**Before this PR** The older logs were like this:
```
E0426 00:06:33.763347 21247 kubelet_volumes.go:114] Orphaned pod "35c4a858-2a12-11e7-910c-42010af00003" found, but volume paths are still present on disk.
E0426 00:06:33.763400 21247 kubelet_volumes.go:114] Orphaned pod "e7676365-1580-11e7-8c27-42010af00003" found, but volume paths are still present on disk.
```
The problem being that, all the volumes were spammed w/ no summary info.
**After this PR** the logs look like this:
```
E0426 01:32:27.295568 22261 kubelet_volumes.go:129] Orphaned pod "408b060e-2a1d-11e7-90e8-42010af00003" found, but volume paths are still present on disk. : There were a total of 2 errors similar to this. Turn up verbosity to see them.
E0426 01:32:29.295515 22261 kubelet_volumes.go:129] Orphaned pod "408b060e-2a1d-11e7-90e8-42010af00003" found, but volume paths are still present on disk. : There were a total of 2 errors similar to this. Turn up verbosity to see them.
E0426 01:32:31.293180 22261 kubelet_volumes.go:129] Orphaned pod "408b060e-2a1d-11e7-90e8-42010af00003" found, but volume paths are still present on disk. : There were a total of 2 errors similar to this. Turn up verbosity to see them.
```
And with logging turned up, the extra info logs are shown with details:
```
E0426 01:34:21.933983 26010 kubelet_volumes.go:129] Orphaned pod "1c565800-2a20-11e7-bbc2-42010af00003" found, but volume paths are still present on disk. : There were a total of 3 errors similar to this. Turn up verbosity to see them.
I0426 01:34:21.934010 26010 kubelet_volumes.go:131] Orphan pod: Orphaned pod "1c565800-2a20-11e7-bbc2-42010af00003" found, but volume paths are still present on disk.
I0426 01:34:21.934015 26010 kubelet_volumes.go:131] Orphan pod: Orphaned pod "408b060e-2a1d-11e7-90e8-42010af00003" found, but volume paths are still present on disk.
I0426 01:34:21.934019 26010 kubelet_volumes.go:131] Orphan pod: Orphaned pod "e7676365-1580-11e7-8c27-42010af00003" found, but volume paths are still present on disk.
```
**Release note**
```release-note
Roll up volume error messages in the kubelet sync loop.
```
Several of these loops overlap, and when they are the reason a failure
is happening it is difficult to sort them out. Slighly misalign these
loops to make their impact obvious.
Automatic merge from submit-queue
don't HandleError on container start failure
Failing to start containers is a common error case if there is something wrong with the container image or environment like missing mounts/configs/permissions/etc. Not only is it common; it is reoccurring as backoff happens and new attempts to start the container are made. `HandleError` it too verbose for this very common situation.
Replace `HandleError` with `glog.V(3).Infof`
xref https://github.com/openshift/origin/issues/13889
@smarterclayton @derekwaynecarr @eparis
Automatic merge from submit-queue (batch tested with PRs 45052, 44983, 41254)
Non-controversial part of #44523
For easier review of #44523, i extracted the non-controversial part out to this PR.
Automatic merge from submit-queue (batch tested with PRs 42740, 44980, 45039, 41627, 45044)
Improved code coverage for /pkg/kubelet/types
**What this PR does / why we need it**:
The test coverage for /pkg/kubelet/types was increased from 50% to 87.5%
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
```
Automatic merge from submit-queue (batch tested with PRs 44970, 43618)
CRI: Fix StopContainer timeout
Fixes https://github.com/kubernetes/kubernetes/issues/44956.
I verified this PR with the example provided in https://github.com/kubernetes/kubernetes/issues/44956, and now pod deletion will respect grace period timeout:
```
NAME READY STATUS RESTARTS AGE
gracefully-terminating-pod 1/1 Terminating 0 6m
```
@dchen1107 @yujuhong @feiskyer /cc @kubernetes/sig-node-bugs
Add support for following redirects to the SpdyRoundTripper. This is
necessary for clients using it directly (e.g. the apiserver talking
directly to the kubelet) because the CRI streaming server issues a
redirect for streaming requests.
Also extract common logic for following redirects.
Automatic merge from submit-queue
Add bootstrap support to certificate manager.
Adds configuration options to certificate manager for using bootstrap cert/key
pairs to handle the scenario where new nodes are initialized using a generic
cert/key pair. Bootstrap cert/key pairs are quickly rotated, independent of
duration remaining, so that each kubelet has a unique cert/key pair.
Automatic merge from submit-queue (batch tested with PRs 42202, 40784, 44642, 44623, 44761)
fix comment error for network plugin
**What this PR does / why we need it**:
**Which issue this PR fixes** : fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
** reason for this change **
CNI has recently introduced a new configuration list feature. This
allows for plugin chaining. It also supports varied plugin versions.
Automatic merge from submit-queue (batch tested with PRs 41849, 42033)
fix TODO: find and add active pods for dswp
loops through the list of active pods and ensures that each one exists in the desired state of the world cache
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 44469, 44566, 44467, 44526)
Kubelet:rkt Fix the hostPath Volume creation
**What this PR does / why we need it**:
This PR fix the `hostPath` volume when the path exist and it's not a directory.
At the moment, the creation of a `hostPath` volume for an existing file leads to this error:
> kubelet[1984]: E0413 07:53:16.480922 1984 pod_workers.go:184] Error syncing pod 38359a57-1fb1-11e7-a484-76870fe7db83, skipping: failed to SyncPod: mkdir /usr/share/coreos/lsb-release: not a directory
**Special notes for your reviewer**:
You can have a look to the difference with this [gist](https://gist.github.com/JulienBalestra/28ae15efc8a1393d350300880c07ff4f)
Automatic merge from submit-queue
comment spelling correction in custommetrics
**What this PR does / why we need it**: fix spelling in a comment
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: fixes #
**Special notes for your reviewer**:
**Release note**:
```release-note
```
Automatic merge from submit-queue (batch tested with PRs 40055, 42085, 44509, 44568, 43956)
Fix gofmt errors
**What this PR does / why we need it**:
There were some gofmt errors on master. Ran the following to fix:
```
hack/verify-gofmt.sh | grep ^diff | awk '{ print $2 }' | xargs gofmt -w -s
```
**Which issue this PR fixes** *(optional, in `fixes #<issue number>(, fixes #<issue_number>, ...)` format, will close that issue when PR gets merged)*: none
**Special notes for your reviewer**:
**Release note**:
```release-note
NONE
```
Automatic merge from submit-queue (batch tested with PRs 44569, 44398)
Move v1/refs and v1/resource
This PR moves pkg/api/v1/ref.go and pkg/api/v1/resource_helper.go to their own sub packages, it's very similar to 44299 and 44302.
The PR is mostly mechanical, except that
* i moved some utility function from resource.go to pkg/api/v1/pod and pkg/api/v1/node, as they are more appropriate
* i updated the staging/copy.sh to copy the new subpackages, so that helper functions are copied. We can get rid of this copy after client-go stops copying API types.
rktnetes is not a CRI implementation, and does not provide runtime
conditions. This change fixes the issue where rkt will never be
considered running from kubelet's point of view.
Automatic merge from submit-queue (batch tested with PRs 44364, 44361, 42498)
Fix the certificate rotation threshold and add jitter.
Adjusts the certificate rotation threshold to be fixed, with some jitter to
spread out the load on the Certificate Signing Request API. The rotation
threshold is fixed at 20% now, meaning when 20% of the certificate's total
duration is remaining, the certificate manager will attempt to rotate, with
jitter +/-10%. For certificates of duration 1 month that means they will
rotate after 24 days, +/- 3 days.
On a 6000 node cluster, assuming all nodes added at nearly the same time, this
should result in 6000 nodes rotating spread over 6 days (total range of the
jitter), or ~42 nodes / hour requesting new certificates.
Automatic merge from submit-queue (batch tested with PRs 44406, 41543, 44071, 44374, 44299)
Decouple remotecommand
Refactored unversioned/remotecommand to decouple it from undesirable dependencies:
- term package now is not required, and functionality required to resize terminal size can be plugged in directly in kubectl
- in order to remove dependency on kubelet package - constants from kubelet/server/remotecommand were moved to separate util package (pkg/util/remotecommand)
- remotecommand_test.go moved to pkg/client/tests module
Automatic merge from submit-queue
CRI: Stop following container log when container exited.
Fixes https://github.com/kubernetes/kubernetes/issues/44340.
This PR changed kubelet to periodically check whether container is running when following container logs, and stop following when container exited.
I've tried this PR in my local cluster:
```
Wed Apr 12 20:23:54 UTC 2017
Wed Apr 12 20:23:58 UTC 2017
Wed Apr 12 20:24:02 UTC 2017
Wed Apr 12 20:24:06 UTC 2017
Wed Apr 12 20:24:10 UTC 2017
Wed Apr 12 20:24:14 UTC 2017
Wed Apr 12 20:24:18 UTC 2017
Wed Apr 12 20:24:22 UTC 2017
Wed Apr 12 20:24:26 UTC 2017
Wed Apr 12 20:24:30 UTC 2017
Wed Apr 12 20:24:34 UTC 2017
Wed Apr 12 20:24:38 UTC 2017
Wed Apr 12 20:24:42 UTC 2017
Wed Apr 12 20:24:46 UTC 2017
failed to wait logs for log file "/var/log/pods/1d54634c7b31346fc3219f5e0b7507cc/nginx_0.log": container "b9a17a2c53550c3703ab350d85911743af8bf164a41813544fd08fb9585f7501" is not running (state="CONTAINER_EXITED")
```
The only difference is that `ReadLogs` will return error when container exits during following. I'm not sure whether we should get rid of it or not.
@yujuhong @feiskyer @JorritSalverda
/cc @kubernetes/sig-node-bugs
**Release note**:
```release-note
`kubectl logs -f` now stops following when container stops.
```
Automatic merge from submit-queue
Add prometheus metrics for age of stats used for evictions.
Completes #42923
This PR adds metrics for evictions, and records how stale data used for evictions is.
cc @vishh @derekwaynecarr @kubernetes/sig-node-pr-reviews
