-
Bug
-
Resolution: Done-Errata
-
Major
-
None
-
4.16
-
None
-
SDN Sprint 259, SDN Sprint 260
-
2
-
False
-
-
-
Bug Fix
-
Done
-
This is a clone of issue OCPBUGS-39438. The following is the description of the original issue:
—
Description of problem: If a customer applies ethtool configuration to the interface used in br-ex, that configuration will be dropped when br-ex is created. We need to read and apply the configuration from the interface to the phys0 connection profile, as described in https://issues.redhat.com/browse/RHEL-56741?focusedId=25465040&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-25465040
Version-Release number of selected component (if applicable): 4.16
How reproducible: Always
Steps to Reproduce:
1. Deploy a cluster with an NMState config that sets the ethtool.feature.esp-tx-csum-hw-offload field to "off"
2.
3.
Actual results: The ethtool setting is only applied to the interface profile which is disabled after configure-ovs runs
Expected results: The ethtool setting is present on the configure-ovs-created profile
Additional info:
Affected Platforms: VSphere. Probably baremetal too and possibly others.
- blocks
-
OCPBUGS-41256 Configure-ovs doesn't persist ethtool configuration
- Closed
- clones
-
OCPBUGS-39438 Configure-ovs doesn't persist ethtool configuration
- Verified
- is blocked by
-
OCPBUGS-39438 Configure-ovs doesn't persist ethtool configuration
- Verified
- is cloned by
-
OCPBUGS-41256 Configure-ovs doesn't persist ethtool configuration
- Closed
- links to
-
RHBA-2024:7922 OpenShift Container Platform 4.17.z bug fix update