Bug 1885132

Summary:	[OVN] Run OVN tasks on host re-install flow
Product:	[oVirt] ovirt-engine	Reporter:	Michael Burman <mburman>
Component:	BLL.Network	Assignee:	Dominik Holler <dholler>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Michael Burman <mburman>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	4.4.3.5	CC:	bugs, dholler, mperina
Target Milestone:	ovirt-4.4.4	Keywords:	Reopened
Target Release:	4.4.4	Flags:	pm-rhel: ovirt-4.4+
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:	ovirt-engine-4.4.4	Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2020-12-21 12:36:47 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	Network	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Michael Burman 2020-10-05 07:47:28 UTC

Description of problem:
[OVN] Run OVN tasks on host re-install flow

The OVN tasks are not executed as they should during the host re-install flow.
I'm pretty sure this is a regression, i know this area of code was otuched lately.

When performing a host re-install the OVN tasks doesn't triggered during the ansible deploy as they should.

Version-Release number of selected component (if applicable):
rhvm-4.4.3.5-0.5.el8ev.noarch
ovirt-ansible-roles-1.2.3-1.el8ev.noarch
ansible-runner-service-1.0.6-2.el8ev.noarch
ansible-runner-1.4.5-1.el8ar.noarch
ansible-2.9.13-1.el8ae.noarch

How reproducible:
100%

Steps to Reproduce:
1. Add fresh host to a cluster with OVN as network default provider
2. Set host to maintenance
3. Do re-install

Actual results:
1. OVN tasks run as expected and OVN configured
3. OVN tasks doesn't triggered and OVN wasn't configured

Expected results:
3. OVN tasks should be triggered on host re-install as well

Additional info:
I'm pretty sure this is a regression.

Comment 2 Dominik Holler 2020-10-06 16:02:30 UTC

Looks like the association of the host to the network provider is lost sometimes:

engine=# select vds_name,openstack_network_provider_id from vds_static;
   vds_name   |    openstack_network_provider_id     
--------------+--------------------------------------
 host_mixed_1 | 
 host_mixed_2 | a0ee5275-cd53-4884-848c-2a3e2cfe504c
 host_mixed_3 | 

This is an inconsistency in the db, because the network provider is managed completely on cluster level.
One way to fix the bug would be to find the flow which created the inconsistency.
Alternatively we could think about replacing the openstack_network_provider_id column by a reference to the network provider of the cluster and remove all code which writes to this column.

Comment 3 Sandro Bonazzola 2020-11-06 13:46:25 UTC

Can be tested with 4.4.3.

Comment 7 Dominik Holler 2020-11-09 07:53:31 UTC

Sorry for the confusion, the fix is ovirt-engine-4.4.4 only.

Comment 8 Michael Burman 2020-11-15 10:17:28 UTC

Verified on - rhvm-4.4.4-0.1.el8ev.noarch

Comment 9 Sandro Bonazzola 2020-12-21 12:36:47 UTC

This bugzilla is included in oVirt 4.4.4 release, published on December 21st 2020.

Since the problem described in this bug report should be resolved in oVirt 4.4.4 release, it has been closed with a resolution of CURRENT RELEASE.

If the solution does not work for you, please open a new bug report.