Bug 1320411

Summary:	[infrastructure_public_284]Should not schedule pod without request on node which "allocatable" value is 0
Product:	OKD	Reporter:	DeShuai Ma <dma>
Component:	Pod	Assignee:	Andy Goldstein <agoldste>
Status:	CLOSED NOTABUG	QA Contact:	DeShuai Ma <dma>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	3.x	CC:	aos-bugs, decarr, mmccomas
Target Milestone:	---
Target Release:	---
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2016-03-23 13:42:52 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:

Description DeShuai Ma 2016-03-23 07:27:30 UTC

Description of problem:
When pod without request value, it still can be scheduled to node which "allocatable" value is 0.
Even there is another node has enough "allocatable" resouce, it sometime still can be scheduled to the node which "allocatable" value is 0
http://pastebin.test.redhat.com/359139

Version-Release number of selected component (if applicable):
openshift v3.2.0.6
kubernetes v1.2.0-36-g4a3f9c5
etcd 2.2.5

How reproducible:
Always

Steps to Reproduce:
1.Set node reserved to equal node's capacity
kubeletArguments:
  system-reserved:
    - "cpu=1,memory=2G"
  kube-reserved:
    - "cpu=1,memory=2G"

2.Check node allocatable
"allocatable": {
    "cpu": "0",
    "memory": "0",
    "pods": "110"
}

3.Create a pod without request
$ oc create -f https://raw.githubusercontent.com/openshift/origin/master/examples/hello-openshift/hello-pod.json

4.Check pod status

Actual results:
4.It's running and be scheduled to node which cpu and memory is 0.

Expected results:
4.The pod should pending.

Additional info:

Comment 1 Andy Goldstein 2016-03-23 13:42:52 UTC

This is working as designed. Best effort pods (those with no request set) are allowed to land on a node regardless of its allocatable amount. Best effort pods do not cause allocatable to decrease.