Automatic merge from submit-queue
Swap in tests ownership
To make the test ownership more closer to actual area of expertise I made the following swap. I included @mtaufen to close the cycle. Please wait with applying lgtm label for the second reviewer.
Automatic merge from submit-queue
Restore etcd from backup
The new script performs disaster recovery of etcd from backup.
https://github.com/kubernetes/kubernetes/issues/20504
@alex-mohr @lavalamp @kubernetes/goog-gke
Automatic merge from submit-queue
Fixes for golint tip
golint as of
3390df4df2/lint.go (L1440-L1442)
requires that any function that has a context.Context argument have said
argument in the first position. This commit fixes the one function we had where
it wasn't.
cc @timothysc @wojtek-t @smarterclayton @jessfraz @kubernetes/rh-cluster-infra @liggitt @sttts @deads2k @eparis
Automatic merge from submit-queue
kubeadm: Stop assuming full ownership of /etc/kubernetes.
Packages may auto-create directories in /etc/kubernetes, and users also
need files such as cloud-config.json to be present and preserved at
their default locations in /etc/kubernetes. As such this modifies
pre-flight checks to only require the absence of the files and
directories we explicitly create in kubeadm.
Reset is similarly modified to not wipe out /etc/kubernetes entirely.
When resetting directories we also now preserve the directory itself,
but delete it's contents.
Also adds tests for reset command logic specifically for /etc/kubernetes
cleanup, to ensure user files are not inadvertently wiped out.
This is doc seems out of place. It is next to the build system for this
addon and is out of date, with del tags blocking out parts. Lets delete
it and I will merge the good parts over to the admin guide:
http://kubernetes.io/docs/admin/dns/
when using kubectl set resources it resets all resource fields that are not being set.
for example
# kubectl set resources deployments nginx --limits=cpu=100m
followed by
# kubectl set resources deployments nginx --limits=memory=256Mi
would result in the nginx deployment only limiting memory at 256Mi with the previous
limit placed on the cpu being wiped out. This behavior is corrected so that each invocation
only modifies fields set in that command and changed the testing so that the desired behavior
is checked.
Also a typo:
you must specify an update to requests or limits or (in the form of --requests/--limits)
corrected to
you must specify an update to requests or limits (in the form of --requests/--limits)
changelog:
- fixed a typo in hack/make-rules/test-cmd.sh "effecting" to "affecting"
Automatic merge from submit-queue
Cadvisor root path configuration
This solves #33444 and reverts PR #33520. This takes the root directory and passes it to cadvisor, which sets up the "nodefs" based on the provided path, rather than using "/" as before.
This PR is pending on changes in cadvisor, and will not pass tests until those changes are merged, and cadvisor godeps are updated.
```release-note
kubelet summary rootfs now refers to the filesystem that contains the Kubelet RootDirectory (var/lib/kubelet) instead of cadvisor's rootfs ( / ), since they may be different filesystems.
```
golint as of
3390df4df2/lint.go (L1440-L1442)
requires that any function that has a context.Context argument have said
argument in the first position. This commit fixes the one function we had where
it wasn't.
This will allow packages to maintain ownership of config and data
directories, which may carry selinux or other attributes that should be
preserved, but we do not wish to manage within kubeadm itself.
