Revert "Scheduler should use shared informer for pods"

[wojtek-t]

This reverts commit 479f01d.

@smarterclayton @deads2k - I don't really have time to debug this today, so for now I'm reverting this commit. We should try to understand what does it change with respect to #46196 (assuming that this will fix the problem).

[wojtek-t]

@k8s-bot unit test this

[gmarek]

/lgtm

[k8s-github-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: gmarek, wojtek-t

Needs approval from an approver in each of these OWNERS Files:

• plugin/cmd/kube-scheduler/OWNERS [wojtek-t]
• plugin/pkg/scheduler/OWNERS [wojtek-t]
• test/OWNERS [gmarek,wojtek-t]

You can indicate your approval by writing /approve in a comment
You can cancel your approval by writing /approve cancel in a comment

[k8s-ci-robot]

@wojtek-t: The following test(s) failed:

Test name	Commit	Details	Rerun command
pull-kubernetes-federation-e2e-gce	d5227a3	link	@k8s-bot pull-kubernetes-federation-e2e-gce test this

Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR.

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here.

[k8s-github-robot]

Automatic merge from submit-queue

[wojtek-t]

@deads2k @smarterclayton - so reverting this fixed the problem. It must be something wrong with this commit, though I don't know what exactly.

[deads2k]

@deads2k @smarterclayton - so reverting this fixed the problem. It must be something wrong with this commit, though I don't know what exactly.

@wojtek-t Thanks for the heads up and for targeting the revert at a commit.

[wojtek-t]

@wojtek-t Thanks for the heads up and for targeting the revert at a commit.

No problem. I was trying to understand the problem, but the commit looks reasonable and I didn't have time for reproducing adding logs, etc. to really debug it. But since this is just one commit, hopefully you will be able to do it.
Note that this also broke some of the scheduler predicates tests (see in the issue for it), which might be easier to reproduce even.

[smarterclayton]

Yeah, hopefully this is obvious on the predicates test.

[smarterclayton]

I0521 05:16:54.615] May 21 05:16:54.558: INFO: At 2017-05-21 05:06:54 -0700 PDT - event for maxp-311: {default-scheduler } FailedScheduling: No nodes are available that match all of the following predicates:: Insufficient pods (3).
I0521 05:16:54.615] May 21 05:16:54.558: INFO: At 2017-05-21 05:06:54 -0700 PDT - event for maxp-312: {default-scheduler } FailedScheduling: No nodes are available that match all of the following predicates:: Insufficient pods (3).
I0521 05:16:54.615] May 21 05:16:54.558: INFO: At 2017-05-21 05:06:54 -0700 PDT - event for maxp-313: {default-scheduler } FailedScheduling: No nodes are available that match all of the following predicates:: Insufficient pods (3).

and

I0521 05:16:54.780] May 21 05:16:54.661: INFO: e2e-image-puller-bootstrap-e2e-minion-group-3029  bootstrap-e2e-minion-group-3029  Succeeded         [{Initialized True 0001-01-01 00:00:00 +0000 UTC 2017-05-21 04:35:24 -0700 PDT PodCompleted } {Ready False 0001-01-01 00:00:00 +0000 UTC 2017-05-21 04:35:28 -0700 PDT PodCompleted } {PodScheduled True 0001-01-01 00:00:00 +0000 UTC 2017-05-21 04:35:24 -0700 PDT  }]
I0521 05:16:54.780] May 21 05:16:54.661: INFO: e2e-image-puller-bootstrap-e2e-minion-group-lnb7  bootstrap-e2e-minion-group-lnb7  Succeeded         [{Initialized True 0001-01-01 00:00:00 +0000 UTC 2017-05-21 04:35:30 -0700 PDT PodCompleted } {Ready False 0001-01-01 00:00:00 +0000 UTC 2017-05-21 04:35:34 -0700 PDT PodCompleted } {PodScheduled True 0001-01-01 00:00:00 +0000 UTC 2017-05-21 04:35:30 -0700 PDT  }]
I0521 05:16:54.780] May 21 05:16:54.661: INFO: e2e-image-puller-bootstrap-e2e-minion-group-zwjz  bootstrap-e2e-minion-group-zwjz  Succeeded         [{Initialized True 0001-01-01 00:00:00 +0000 UTC 2017-05-21 04:35:26 -0700 PDT PodCompleted } {Ready False 0001-01-01 00:00:00 +0000 UTC 2017-05-21 04:35:30 -0700 PDT PodCompleted } {PodScheduled True 0001-01-01 00:00:00 +0000 UTC 2017-05-21 04:35:26 -0700 PDT  }]

https://storage.googleapis.com/kubernetes-jenkins/logs/ci-kubernetes-e2e-gce-serial/1102/build-log.txt

[smarterclayton]

So it looks like I am not properly excluding things from the cache. That would explain the predictability.