Bug 1871819

Summary:	Prepare default OVN configuration for scaling scenarios
Product:	Red Hat Enterprise Virtualization Manager	Reporter:	Dominik Holler <dholler>
Component:	ovirt-provider-ovn	Assignee:	Dominik Holler <dholler>
Status:	CLOSED ERRATA	QA Contact:	Michael Burman <mburman>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	4.3.9	CC:	dfodor, emarcus, mburman, mperina, pelauter
Target Milestone:	ovirt-4.4.3	Keywords:	ZStream
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:	ovirt-engine-4.4.3.5	Doc Type:	Bug Fix
Doc Text:	Previously, when changes were made in the logical network, the ovn-controller on the host sometimes exceeded the timeout interval during recalculation, and calculation was triggered repeatedly. As a result, OVN networking failed. In this release, recalculation by the ovn-controller is only triggered once per change, and OVN networking is maintained.	Story Points:	---
Clone Of:		Environment:
Last Closed:	2020-11-24 13:10:38 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 Dominik Holler 2020-08-24 11:15:44 UTC

Description of problem:
The default parameters of OVN are not suitable for scaling scenarios.

Steps to Reproduce:
1. Configure ovn using ovirt-engine-setup


Actual results:
The parameters are not configure to be suitable suitable for scaling scenarios.

Expected results:
The parameters are configure to be suitable suitable for scaling scenarios.
 

Additional info:
The relevant parameters are
 - inactivity_probe on ovn-central should be 60000
 - external_ids:ovn-remote-probe-interval on ovn-controller should be 60000
 - external_ids:ovn-openflow-probe-interval on ovn-controller should be 60

Comment 1 Dominik Holler 2020-09-08 09:52:46 UTC

Hints for verification:
on Engine machine,
ovn-sbctl get connection . inactivity_probe
should return 60000

on hosts:
ovs-vsctl get open . external_ids:ovn-remote-probe-interval
-> 60000
ovs-vsctl get open . external_ids:ovn-openflow-probe-interval
-> 60

Comment 2 Dominik Holler 2020-09-29 17:40:01 UTC

Fixed in ovirt-engine-ovirt-engine-4.4.3.5 and ovirt-provider-ovn-1.2.32-1.

Comment 4 Michael Burman 2020-10-05 07:51:38 UTC

Verified on - rhvm-4.4.3.5-0.5.el8ev.noarch with:
ovirt-provider-ovn-1.2.32-1.el8ev.noarch
ovirt-provider-ovn-driver-1.2.32-1.el8ev.noarch

Engine:
 ovn-sbctl get connection . inactivity_probe
60000

Host(the modified script is currently only triggered during host install via ansible):
 ovs-vsctl get open . external_ids:ovn-remote-probe-interval
"60000"
ovs-vsctl get open . external_ids:ovn-openflow-probe-interval
"60"

Note: The OVN tasks are not triggered during re-install flow. They are only triggered on install flow. Will open a bug to handle the re-install flow, as the OVN tasks should be triggered as well when performing host re-install.
See BZ 1885132

Comment 11 errata-xmlrpc 2020-11-24 13:10:38 UTC

Since the problem described in this bug report should be
resolved in a recent advisory, it has been closed with a
resolution of ERRATA.

For information on the advisory (RHV Engine and Host Common Packages 4.4.z [ovirt-4.4.3]), and where to find the updated
files, follow the link below.

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

https://access.redhat.com/errata/RHBA-2020:5212