OCPBUGS-113532: ovn-kubernetes: Move MNP from ConfigMap to CLI flags - #3136
OCPBUGS-113532: ovn-kubernetes: Move MNP from ConfigMap to CLI flags#3136raphaelvrosa wants to merge 1 commit into
Conversation
MultiNetworkPolicy is not enforced on UDN secondary interfaces in OCP 5.0 because ovnkube-control-plane pods do not restart when the ConfigMap is updated with enable-multi-networkpolicy=true. Root cause: PR openshift#2944 moved MNP enablement to ConfigMap, but: 1. The ConfigMap hash only includes 008-script-lib.yaml, not 004-config.yaml 2. ovnkube-control-plane has no hash annotation to trigger restarts Solution: Follow the same pattern as commit f4734c5 (OCPBUGS-78731): move MNP back to CLI flags where pod restart happens automatically on spec changes. This creates consistency with multicast (also a CLI flag) and avoids the ConfigMap hash timing issues that caused the original NVIDIA-554 fix to be reverted. How to verify: 1. Deploy cluster with UseMultiNetworkPolicy=false - Verify ovnkube-control-plane pods lack --enable-multi-networkpolicy flag 2. Set UseMultiNetworkPolicy=true via network.operator/cluster - Verify ovnkube-control-plane pods restart with --enable-multi-networkpolicy 3. Test MNP enforcement on UDN Layer2/Layer3 networks 4. Verify DPU-host mode still works correctly 5. CI lanes: e2e-aws-ovn, e2e-gcp-ovn, e2e-metal-ipi-ovn-dualstack Tested: - MNP works on UDN Layer2 and Layer3 networks - Pods restart when UseMultiNetworkPolicy changes - DPU-host mode still works correctly - Other ConfigMap features (egress-ip, etc.) unaffected Jira: OCPBUGS-88063 docs: Clarify CLI flags are for control-plane pod restarts Update ovn_node_mode.md to specify that CLI flags are used for features requiring ovnkube-control-plane pod restarts, not just any pod restarts. This clarifies that ovnkube-node pods already restart when the ConfigMap changes, so CLI flags are only needed for control-plane-specific features. Addresses review feedback from danwinship. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com> (cherry picked from commit 17f08a7)
|
Pipeline controller notification For optional jobs, comment This repository is configured in: LGTM mode |
|
Important Review skippedAuto reviews are disabled on base/target branches other than the default branch. Please check the settings in the CodeRabbit UI or the ⚙️ Run configurationConfiguration used: Repository: openshift/coderabbit/.coderabbit.yaml Review profile: CHILL Plan: Enterprise Run ID: You can disable this status message by setting the Use the checkbox below for a quick retry:
Comment |
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: raphaelvrosa The full list of commands accepted by this bot can be found here. DetailsNeeds approval from an approver in each of these files:Approvers can indicate their approval by writing |
|
@raphaelvrosa: This pull request references Jira Issue OCPBUGS-113532, which is valid. The bug has been moved to the POST state. 7 validation(s) were run on this bug
No GitHub users were found matching the public email listed for the QA contact in Jira (core-networking-bot@redhat.com), skipping review request. The bug has been updated to refer to the pull request using the external bug tracker. DetailsIn response to this:
Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the openshift-eng/jira-lifecycle-plugin repository. |
|
/pipeline required |
|
Scheduling required tests: Scheduling tests matching the |
|
@raphaelvrosa: The following tests failed, say
Full PR test history. Your PR dashboard. DetailsInstructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes-sigs/prow repository. I understand the commands that are listed here. |
MultiNetworkPolicy is not enforced on UDN secondary interfaces in OCP 5.0 because ovnkube-control-plane pods do not restart when the ConfigMap is updated with enable-multi-networkpolicy=true.
Root cause: PR #2944 moved MNP enablement to ConfigMap, but:
Solution: Follow the same pattern as commit f4734c5 (OCPBUGS-78731): move MNP back to CLI flags where pod restart happens automatically on spec changes.
This creates consistency with multicast (also a CLI flag) and avoids the ConfigMap hash timing issues that caused the original NVIDIA-554 fix to be reverted.
How to verify:
Tested:
Jira: OCPBUGS-88063
docs: Clarify CLI flags are for control-plane pod restarts
Update ovn_node_mode.md to specify that CLI flags are used for features requiring ovnkube-control-plane pod restarts, not just any pod restarts. This clarifies that ovnkube-node pods already restart when the ConfigMap changes, so CLI flags are only needed for control-plane-specific features.
Addresses review feedback from danwinship.
(cherry picked from commit 17f08a7)