Move to github.com/munnerz/goautoneg as bitbucket is flaky!
Change-Id: Iaa6e964ef0d6f308eea59bcc6f365ecd7dbf0784
Kubernetes-commit: 16fd72d6c91ba466a0e955a1d59a6c8d9e8791bc
This package contains public/private key utilities copied directly from
client-go/util/cert. All imports were updated.
Future PRs will actually refactor the libraries.
Updates #71004
Kubernetes-commit: 18458392ca24c85c688e655aace1afd04f864cbd
Fixes#71730
0 indicates standby, 1 indicates master, label indicates which lease.
Tweaked name and documentation
Factored in Mike Danese feedback.
Removed dependency on prometheus from client-go using adapter.
Centralized adapter import.
Fixed godeps
Fixed boilerplate.
Put in fixes for caesarxuchao
Kubernetes-commit: f1926573804ed2c172c91d1022203d0699210138
Clean up the code paths that lead to objects being transformed and output with negotiation.
Remove some duplicate code that was not consistent. Now, watch will respond correctly to
Table and PartialObjectMetadata requests. Add unit and integration tests.
When transforming responses to Tables, only the first watch event for a given type will
include the columns. Columns will not change unless the watch is restarted.
Add a volume attachment printer and tighten up table validation error cases.
Disable protobuf from table conversion because Tables don't have protobuf because they
use `interface{}`
Kubernetes-commit: 3230a0b4fd14a6166f8362d4732e199e8779c426
The current code simply exits without continuing to renew the lease, which means
participants using a slower lease duration might have to wait multiple minutes
before a new leader is elected. Allow an optional flag to be set on
LeaderElectionConfig that will release the lease when the calling context is
cancelled. Callers *must* ensure their lease guarded code has completed before
the context is cancelled, or other processes may acquire the lease before this
lease has released.
Add an example command that demonstrates how cancellation could be done.
As a convenience to users, make event recorder optional - not all users of the
lock code will need a recorder.
Kubernetes-commit: 09890b6c48da8e85237a5674d6256900f482b0a5
This commit fixes a bug in the client-go/testing fixture whereby a
watcher would fire twice for objects with no namespace.
Kubernetes-commit: bd268c9971efd4db46074a80e52ce765583e0787
This reverts commit 0af19875add7deb562b2cf7bf6b1d273c44bab1b.
Revert "Ensure the bootstrap rotation code is tested by forcing rotation"
This reverts commit de293b2d7ddb687850258370f2a7f30f224f0ec1.
Kubernetes-commit: 34642222676640b3c1dd255cc453000f2743ccde
Expose both a Stop() method (for cleanup) and a method to force
cert rotation, but only expose Stop() on the interface.
Verify that we choose the correct client.
Kubernetes-commit: de293b2d7ddb687850258370f2a7f30f224f0ec1
Not every object kind can be registered via tracker .Add() called as
part of SimpleClientset initialization. This is because .Add() relies
on UnsafeGuessKindToResource to convert object kinds into resource
type names, which is broken for some resources. An example of an
affected kind is NetworkAttachmentDefinitions CRD that uses
network-attachment-definitions as its resource type name. When
UnsafeGuessKindToResource is called for this kind, it returns
networkattachmentdefinitions (without dashes).
As per the comment inside .Add, kinds affected by
UnsafeGuessKindToResource deficiencies should instead register objects
using tracker .Create() method. Problem is, current SimpleClientset
struct definition doesn't expose the object tracker in any way, which
makes it impossible to properly register these kinds at all.
To address the issue, this change modifies the definition of
SimpleClientset struct to expose the object tracker used via Tracker()
method.
Kubernetes-commit: d68cd8a0c7e6137ca4219078a3d651ecff03c21f
