Bug 1957526

Summary:	[RFE] Auto Pinning Policy 'Existing' to keep the existing vNUMA count of the VM, instead of matching the host
Product:	[oVirt] ovirt-engine	Reporter:	Germano Veit Michel <gveitmic>
Component:	BLL.Virt	Assignee:	Liran Rotenberg <lrotenbe>
Status:	CLOSED WONTFIX	QA Contact:	Polina <pagranat>
Severity:	medium	Docs Contact:
Priority:	unspecified
Version:	4.4.6	CC:	ahadas, bugs, sigbjorn.lie
Target Milestone:	---	Keywords:	FutureFeature
Target Release:	---	Flags:	pm-rhel: ovirt-4.5? pm-rhel: planning_ack? ahadas: devel_ack? pm-rhel: testing_ack?
Hardware:	Unspecified
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2022-04-06 15:41:31 UTC	Type:	Bug
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	Virt	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Germano Veit Michel 2021-05-05 23:57:58 UTC

Description of problem:

The 'Existing' policy, which is default for HP VMs, keeps the vCPU count of the VM but it adjusts the vNUMA count of the VM to match the host. This is not optimal on some cases (i.e. 4 NUMA host but the VM fits perfectly fine on 2, or 2 NUMAs on the host but the VM is fine on 1). It is splitting the VM unnecessarily over more NUMA nodes.

So this RFE is to make the 'Existing' policy keep both the user configured vCPU count (which it already does) and vNUMA. So the behaviour is more coherent and topology is really unchanged.

https://bugzilla.redhat.com/show_bug.cgi?id=1954878#c7

Comment 1 Sandro Bonazzola 2022-03-29 16:16:40 UTC

We are past 4.5.0 feature freeze, please re-target.

Comment 3 Arik 2022-04-06 15:41:31 UTC

The 'EXISTING' policy has been removed
This new 'DEDICATED' policy respects the vNUMA count when specified - it's not a 1:1 replacement for the 'EXISTING' policy but should be used instead