Bug 1979231

Summary: [ovn] Set inactivity_probe for NB/SB databases
Product: Red Hat OpenStack Reporter: Daniel Alvarez Sanchez <dalvarez>
Component: python-networking-ovnAssignee: Rodolfo Alonso <ralonsoh>
Status: CLOSED ERRATA QA Contact: Eran Kuris <ekuris>
Severity: high Docs Contact:
Priority: medium    
Version: 16.1 (Train)CC: apevec, jlibosva, lhh, lmartins, majopela, scohen
Target Milestone: z8Keywords: Triaged
Target Release: 16.1 (Train on RHEL 8.2)   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: python-networking-ovn-7.3.1-1.20211224103345.4e24f4c.el8ost Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2022-03-24 11:00:12 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:

Description Daniel Alvarez Sanchez 2021-07-05 10:06:51 UTC
The 'inactivity_probe' parameter in OVN dbs schemas [0][1] represents the maximum number of milliseconds of idle time on connection to client before sending an inactivity probe message.

This parameter is never set by Neutron/TripleO and it is important to allow its configuration since it can be crucial at scale or under heavy load to avoid reconnections.

Devstack sets it at deployment time [2] so I'm open to mimic this and move the BZ to TripleO if that's a better approach.


[0] https://github.com/ovn-org/ovn/blob/95e41218fd7ea9c288c95744e80ba7caa90b30e2/ovn-sb.ovsschema#L275
[1] https://github.com/ovn-org/ovn/blob/95e41218fd7ea9c288c95744e80ba7caa90b30e2/ovn-nb.ovsschema#L456
[2] https://opendev.org/openstack/devstack/src/branch/master/lib/neutron_plugins/ovn_agent#L699

Comment 1 Daniel Alvarez Sanchez 2021-07-05 10:12:23 UTC
As a side note, I reported this against Neutron because there might be a bug in ovsdb-server that deletes this inactivity_probe value and it would be a good workaround if we'd reset it from Neutron upon a reconnection.

However, since Neutron is not the only SB client (eg. ovn-controller and metadata-agent), it feels weird to let Neutron take the responsibility of configuring a system wide parameter. IMO, the best approach would be to let the deployer do it based on a cluster configuration.

Comment 9 errata-xmlrpc 2022-03-24 11:00:12 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 (Red Hat OpenStack Platform 16.1.8 bug fix and enhancement advisory), 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-2022:0986