Bug 1889701

Summary:	Network Operator CRD updated in openshift/api is not copied to release payload.
Product:	OpenShift Container Platform	Reporter:	Venkata Siva Teja Areti <vareti>
Component:	Networking	Assignee:	Ricardo Carrillo Cruz <ricarril>
Networking sub component:	openshift-sdn	QA Contact:	zhaozhanqi <zzhao>
Status:	CLOSED NEXTRELEASE	Docs Contact:
Severity:	high
Priority:	high	CC:	aconstan, bbennett, sttts
Version:	4.7	Keywords:	UpcomingSprint
Target Milestone:	---
Target Release:	4.7.0
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2020-12-04 18:32:27 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:
Embargoed:

Description Venkata Siva Teja Areti 2020-10-20 12:16:05 UTC

Description of problem:
Network Operator CRD in openshift/api is not copied to release payload. But instead another CRD present in cluster-network-operator repo is copied. This is causing a mismatch in CRD spec.

For example, log level validation in CRD spec is not seen in a 4.7 cluster. 

> $ oc get crd networks.operator.openshift.io -o jsonpath="{.spec.versions[0].schema.openAPIV3Schema.properties.spec.properties.logLevel}"
> map[description:logLevel allows configuring the logging level of the components deployed by the operator. Currently only Kuryr SDN is affected by this setting. Please note that turning on extensive logging may affect performance. The default value is "Normal". type:string]

https://github.com/openshift/api/blob/master/operator/v1/0000_70_cluster-network-operator_01_crd.yaml#L372-L386

As I understand, this file is copied to the cluster.

https://github.com/openshift/cluster-network-operator/blob/master/manifests/0000_70_cluster-network-operator_01_crd.yaml

Comment 1 Stefan Schimanski 2020-10-20 12:33:58 UTC

Background: openshift/api is the source of thruth for our API definitions. Every operator is supposed to copy from there into their manifests/ directory DURING image builds.

Comment 2 Ricardo Carrillo Cruz 2020-11-11 12:26:55 UTC

will look into this today

Comment 3 Ricardo Carrillo Cruz 2020-12-04 18:32:27 UTC

https://github.com/openshift/cluster-network-operator/pull/863 should have fixed this.