Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 1986406

Summary:	[QE] controller replacement does not remove compute services before the node is deleted
Product:	Red Hat OpenStack	Reporter:	Luigi Tamagnone <ltamagno>
Component:	documentation	Assignee:	Irina <igallagh>
Status:	CLOSED CURRENTRELEASE	QA Contact:	RHOS Documentation Team <rhos-docs>
Severity:	high	Docs Contact:
Priority:	high
Version:	16.1 (Train)	CC:	alifshit, amodi, aschultz, dasmith, dmacpher, eglynn, igallagh, jhakimra, jkreger, jpretori, kchamart, lbezdick, ltamagno, nova-maint, pbabbar, rdiwakar, sbauza, sgordon, smooney, vromanso
Target Milestone:	---	Keywords:	Triaged
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:	1977667
Clones:	1990034 (view as bug list)		Environment:
Last Closed:	2021-12-08 09:47:23 UTC	Type:	---
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:	1990034

Description Luigi Tamagnone 2021-07-27 13:23:46 UTC

Description of problem:
After FFU from 13 to 16.1.
If you try to replace controller nodes with new hardware. 

According to step 17.7 in https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/16.1/html/director_installation_and_usage/replacing-controller-nodes

it requires to remove old services con controller nodes. But it failed on nova-compute

Version-Release number of selected component (if applicable):
RHOSP16.1


Actual results:
openstack compute service list --host <controller-host-name>
failed on: Unexpected API Error
 
Expected results:
openstack compute service list --host <controller-host-name>
works

Additional info:

Comment 9 Artom Lifshitz 2021-08-04 17:05:01 UTC

Because this is getting a bit confusing, here's a summary that I'll reproduce in the other affected BZs.

BZ 1986406 (this BZ): Documentation for controller replacement

BZ 1990034: Documentation for FFU with Ironic as the virt driver

BZ 1977667: Nova fix to allow service deletion when a service has no associated compute nodes.

Comment 13 Artom Lifshitz 2021-09-17 13:36:34 UTC

@Luigi The last few comments in BZ 1977667 should have everything you need, both a hotfix build of the Nova patch, and a script to delete old services in case they don't want to update their deployment to latest z-stream before applying the hotfix. Let's continue in there if there are further issues.

Comment 19 Artom Lifshitz 2021-09-24 13:12:16 UTC

To use the example from [1], specifically the `MariaDB [nova]> select * from services where version != 40;` command line, we see that the "real" services are on version 40, but they have a bunch of old services with version 30 that did not get removed. So they can run the script with a <older_than_version> anywhere between 31 and 39. That will pick up all their old version 30 services, but not their new version 40 services. Does that make sense?

[1] https://bugzilla.redhat.com/show_bug.cgi?id=1977667#c1

Comment 21 Artom Lifshitz 2021-09-24 13:26:45 UTC

Interesting, in that case we should back up a bit and revisit what's the original problem that we're trying to solve is.

Comment 37 Artom Lifshitz 2021-10-28 13:47:43 UTC

*** Bug 2013329 has been marked as a duplicate of this bug. ***

Comment 43 Irina 2021-12-08 09:47:23 UTC

Updated content available on the Customer Portal:

https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/16.2/html/director_installation_and_usage/replacing-controller-nodes#preparing-for-controller-replacement

https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/16.1/html/director_installation_and_usage/replacing-controller-nodes#preparing-for-controller-replacement