Bug 1885848

Summary:	Network operator stuck on message "the cluster operator network has not yet successfully rolled out" while downgrading from 4.6 -> 4.5
Product:	OpenShift Container Platform	Reporter:	pmali
Component:	Networking	Assignee:	Aniket Bhat <anbhat>
Networking sub component:	openshift-sdn	QA Contact:	zhaozhanqi <zzhao>
Status:	CLOSED ERRATA	Docs Contact:
Severity:	medium
Priority:	unspecified	CC:	bbennett, mifiedle, rbrattai
Version:	4.5	Keywords:	TestBlocker
Target Milestone:	---
Target Release:	4.6.0
Hardware:	All
OS:	All
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:
Clones:	1886148 (view as bug list)		Environment:
Last Closed:	2020-10-08 16:47:25 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:
Bug Depends On:
Bug Blocks:	1886127, 1886148

Description pmali 2020-10-07 07:27:14 UTC

Description of problem:

Downgrade(4.6.0-0.nightly-2020-10-05-234751 -> 4.5.0-0.nightly-2020-10-05-204452) stuck on the network operator.

Version-Release number of selected component (if applicable):
4.6.0-0.nightly-2020-10-05-234751
4.5.0-0.nightly-2020-10-05-204452

How reproducible:
Always

Steps to Reproduce:
1.Install 4.6 cluster 
2.Downgrade 4.6 to 4.5 cluster version
3.

Actual results:

Downgrade stuck on the network operator 

Expected results:
Downgrade should successful.

Additional info:

$ oc logs ovs-6xdx4
ovsdb-server: /var/run/openvswitch/ovsdb-server.pid: pidfile check failed (No such process), aborting
Starting ovsdb-server ... failed!

Comment 8 Ben Bennett 2020-10-07 17:59:20 UTC

In 4.6 we don't need to do anything.  We need to handle it on the downgrade.

Comment 10 Ben Bennett 2020-10-08 16:47:25 UTC

This is not something we can fix in 4.6.  We have the 4.5.z PR to handle the case open (https://github.com/openshift/cluster-network-operator/pull/830), and we have a bz open against the MCO to get the ovs disabled on the host.

Note, that only PR https://github.com/openshift/cluster-network-operator/pull/830 is needed to allow downgrade.