Bug 1955029

Summary:	[ovn][migration][docs] MTU considerations during migration
Product:	Red Hat OpenStack	Reporter:	Daniel Alvarez Sanchez <dalvarez>
Component:	documentation	Assignee:	James Smith <jamsmith>
Status:	ON_DEV ---	QA Contact:	Roman Safronov <rsafrono>
Severity:	unspecified	Docs Contact:
Priority:	high
Version:	16.1 (Train)	CC:	averdagu, ekuris, jamsmith, jlibosva, mtomaska
Target Milestone:	z4	Keywords:	Triaged
Target Release:	16.2 (Train on RHEL 8.4)	Flags:	jamsmith: needinfo? (dalvarez) mtomaska: needinfo-
Hardware:	Unspecified
OS:	Unspecified
Whiteboard:
Fixed In Version:		Doc Type:	If docs needed, set a value
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:		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:

Description Daniel Alvarez Sanchez 2021-04-29 09:51:32 UTC

The current migration process assumes moving to Geneve overlay which implies an adjustment of the workloads MTU before the actual migration happens (default MTU for VXLAN is 1450 while default MTU for Geneve is 1442).

Adjusting the MTU is not going to be always possible since it may depend on the actual guest DHCP client, and sometimes the instances won't even have DHCP enabled and may require a hard reboot.

The doc should capture this and warn the user so that, whenever possible, the MTU is left untouched (eg. 1450) if the underlying physical networks would be on Jumbo frames (or an MTU larger than 1508 bytes). This would save the MTU adjustment step and reduce risks.

Comment 2 James Smith 2021-05-04 13:07:38 UTC

SME Daniel
Affects 16.1 and later