rebase 1.9.0 beta.1

[deads2k]

Green:

1. networking
2. verify
3. cross
4. integration
5. end-to-end with router removed
6. unit
7. test-cmd
8. extended builds
9. extended image ecosystem

Current counts:

1. GCE - 5 failures - 4 networking failures
2. extended conformance - 2 failures
3. end-to-end router tests - CI can't run e2e tests anymore because its trying to use golang 1.8.3. I removed them

e2e failures on gce:

• [Feature:DeploymentConfig] deploymentconfigs when run iteratively [Conformance] should only deploy the last deployment [Suite:openshift/conformance/parallel] - flaking
• [sig-network] Networking Granular Checks: Pods should function for intra-pod communication: http [Conformance] [Suite:openshift/conformance/parallel] [Suite:k8s] 5m27s
• [sig-network] Networking Granular Checks: Pods should function for intra-pod communication: udp [Conformance] [Suite:openshift/conformance/parallel] [Suite:k8s] 5m43s
• [sig-network] Networking Granular Checks: Pods should function for node-pod communication: http [Conformance] [Suite:openshift/conformance/parallel] [Suite:k8s] 5m50s
• [sig-network] Networking Granular Checks: Pods should function for node-pod communication: udp [Conformance] [Suite:openshift/conformance/parallel] [Suite:k8s] 5m49s

e2e failures on install:

• [Conformance][Area:Networking][Feature:Router] openshift router metrics The HAProxy router should expose prometheus metrics for a route [Suite:openshift/conformance/parallel] 21s @deads2k reproduced locally
• [Conformance][Area:Networking][Feature:Router] openshift router metrics The HAProxy router should expose the profiling endpoints [Suite:openshift/conformance/parallel] 23s @deads2k reproduced locally

flaky failures

1. [Feature:Builds] Optimized image builds should succeed [Conformance] [Suite:openshift/conformance/parallel] 23s
2. [Feature:Builds][Conformance] build without output image building from templates should create an image from a docker template without an output image reference defined [Suite:openshift/conformance/parallel] 1m5s
3. [sig-storage] Downward API volume should update labels on modification [Conformance] [Suite:openshift/conformance/parallel] [Suite:k8s] 2m19s I think this flakes on master
4. [Feature:DeploymentConfig] deploymentconfigs when run iteratively [Conformance] should only deploy the last deployment [Suite:openshift/conformance/parallel] - flaking

Post-rebase tasks

• dropskip scheduler configz error once we run it in a separate process - dropskip scheduler configz error once we run it in a separate process #17795
• drop skip controller metric error once we run it in a separate process - drop skip controller metric error once we run it in a separate process #17796
• update openshift roles to use role aggregation - update openshift to use role aggregation #17794
• deal with PrintSuccess in a better way PrintSuccess should be used through the cmdutil.Factory #17767
• collapse onto the upstream lease_endpoint_reconciler.go - collapse onto the upstream lease_endpoint_reconciler.go #17797
• make a quota union registry - make a quota union registry #17802
• update quota to auto-ignore resources that aren't gettable or deletable - update quota to auto-ignore resources that aren't gettable or deletable #17801
• replenishment controller for origin evaluators, possibly generic for namespace - replenishment controller for origin evaluators, possibly generic for namespace #17803
• update all admission plugins to new interfaces - update openshift admission to have mutators and validator #17792
• add our types to the kubectl scheme and drop UPSTREAM: <carry>: switch apply to use the legacyscheme so our types can be handled - Add our types to the kubectl scheme #17793
• re-enable the end-to-end router tests that require a golang 1.9 node - re-enable router integration tests #17755
• sross: update idling to use ownerrefs - Update the idler to use owner refs #17798
• remove DelegatingScaleNamespacer - remove DelegatingScaleNamespacer #17799
• weird storage problems for IntializerConfigurations (alpha), volumeattachments (alpha), and event.events.k8s.io (this one is a validation problem on the server we need to fight through) - etcd storage test failures #17804
• crio constants are inconsistent. There are pulls open to kube upstream. - crio constants are inconsistent (and perhaps wrong) #17800
• oc explain is failing on openshift resource. I suspect the switch from swagger to openapi upstream. oc explain fails with openshift resources #17766
• double check that webhook registration works proxied through the aggregator @sttts WIP: admission/webhook: do not override set client values kubernetes/kubernetes#56758 - pick kubernetes/kubernetes#56758 #17791