Note: This bug is displayed in read-only format because the product is no longer active in Red Hat Bugzilla.

Bug 2063696

Summary:	The procedure in "16.3. Removing Compute nodes" of "Director Installation and Usage" is difficult to understand.
Product:	Red Hat OpenStack	Reporter:	Ryo Hayakawa <rhayakaw>
Component:	documentation	Assignee:	Irina <igallagh>
Status:	CLOSED CURRENTRELEASE	QA Contact:	RHOS Documentation Team <rhos-docs>
Severity:	medium	Docs Contact:
Priority:	unspecified
Version:	16.2 (Train)	CC:	alifshit, astupnik, bdobreli, igallagh
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:		Environment:
Last Closed:	2022-06-07 14:04:41 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:	2023628

Comment 7 Artom Lifshitz 2022-03-23 17:55:49 UTC

I think the 'Removing Compute nodes' is fine as is, with the caveat that we're assuming that step 7 `overcloud node delete` also deletes the nova-compute service entry (which Alex is confirming to be the case in comment #4 point 1). Deleting the nova-compute service will also delete any ComputeNode records, and placement resource providers for that service.

The problematic parts are then:

* The IMPORTANT note in step 8 ii. This should be removed - adding a new compute service, whether it has the same host name or not, is completely independent.

* Section 'Redeploying the Compute service with the same host name' doesn't need to exist. Scaling up with a new compute host that happens to have the same host name as one that was removed previously follows the exact same steps. The new nova-compute service will self-register and create its resource provider entries.

Comment 8 Alex Stupnikov 2022-03-24 15:59:47 UTC

I think that any change to address this bug should be made while keeping in mind a set of fixes for bug #2023628 (thank you Irina for bringing this up): proposed patches for bug #2023628 will change our message and recommendations when it comes to replacing a removed compute node. At the same time I also agree with Artom: separate section to remove compute from a list of placement providers and re-enable nova-compute service is likely useless and should be removed.

I also (like Artom) think that we need to take a second look at IMPORTANT note in step 8 ii, but I rather think that we need to update our message here to address bug #2066541 and ensure that removed compute node's nova-compute service is removed no matter what...

Maybe a smart decision would be to wait till bug #2023628 is addressed and change is published and then take a second look at it and propose some change to address bug #2063696 (current bug) and bug #2066541 (reported by Nozomi).

Comment 22 Irina 2022-06-07 14:04:41 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/assembly_scaling-overcloud-nodes#proc_removing-or-replacing-a-compute-node_scaling-overcloud-nodes

https://access.redhat.com/documentation/en-us/red_hat_openstack_platform/16.1/html/director_installation_and_usage/assembly_scaling-overcloud-nodes#proc_removing-or-replacing-a-compute-node_scaling-overcloud-nodes