SDN-4919,OCPBUGS-38653,OCPBUGS-38267,OCPBUGS-38693: Downstream Merge 14th August 2024 #2265
Conversation
Signed-off-by: arkadeepsen <[email protected]>
Signed-off-by: arkadeepsen <[email protected]>
Bumps [github.com/docker/docker](https://github.com/docker/docker) from 24.0.9+incompatible to 25.0.6+incompatible. - [Release notes](https://github.com/docker/docker/releases) - [Commits](moby/moby@v24.0.9...v25.0.6) --- updated-dependencies: - dependency-name: github.com/docker/docker dependency-type: direct:production ... Signed-off-by: dependabot[bot] <[email protected]>
The issue was a race where the hybrid overlay node was being updated to remove the windows label for testing. However, the update action itself was with a blank original copy of the node which would overwrite l3 gateway config and other OVNK annotations with empty values, causing a bunch of errors. This changes the code to just patch and remove the labels, in order to not corrupt any of the other aspects of the node object itself. Fixes: #4387 Signed-off-by: Tim Rozet <[email protected]>
Reduce linter time by mounting the golangci-lint local cache. Signed-off-by: Or Mergi <[email protected]>
Add support for DNSNameResolver in KIND cluster
This PR does following: - Removes following linux resources if masquerade subnet gets changed (node side): * Removes old V4HostMasqueradeIP and V6HostMasqueradeIP from bridge. * Removes stale neighbour entries V4OVNMasqueradeIP, V6OVNMasqueradeIP, V4DummyNextHopMasqueradeIP and V6DummyNextHopMasqueradeIP if exists. * Removes stale masquerade route added by addMasqueradeRoute() function while starting up the gateway. * Removes stale iptables rules created for masquerade subnet based on ipForwarding and Gateway mode. - Removes following linux resources if masquerade subnet gets changed (ovnkube-controller to NBDB side): * Removes logical router static route used by gateway router and referencing old masquerade subnet. * Removes static mac binding for gateway router's rtoe logical port referencing old masquerade subnet. Note, the node now sets an annotation to indicate its masquerade subnet that it last configured. The node uses this at start up to determine if there has been a change and cleanup is needed. On the ovnkube-controller side, it also uses this annotation to determine if the node has changed. However, it may be racy to rely on this as the node thread may have already updated the annotation by the time the ovnkube-controller side handles the cleanup. Therefore, in addition to the annotation ovnkube-controller will additionally scan for stale routes in NBDB and then derive the route and mac binding to remove that way. In order to facilitate this, the masquerade route now has an external_id present (same as the key used in the annotation) to distinguish which routes are masquerade routes. Failure to delete things is not usually an overall failure for OVNK. Therefore upon failing to clean something up, the error is logged, but startup continues. Finally, kind.sh is updated to use a larger masquerade subnet by default. OVN-Kubernetes defaults themselves remain unchanged. Helm has also been updated to use a larger subnet. Co-authored-by: Tim Rozet <[email protected]> Signed-off-by: Arnab Ghosh <[email protected]>
Support masquerade subnet to be custom configurable as a day2 operation
Signed-off-by: Shane Utt <[email protected]>
fix: link to kind usage docs in contrib README.md
Signed-off-by: Martin Kennelly <[email protected]>
Signed-off-by: Martin Kennelly <[email protected]>
Signed-off-by: Martin Kennelly <[email protected]>
Signed-off-by: Martin Kennelly <[email protected]>
Add test cases to cover this scenario. Signed-off-by: Martin Kennelly <[email protected]>
In order to do that, factor out the code that allocates join IPs and creates the join switch and move it to the base network controller. A follow up commit will move both the `createJoinSwitch` and `createClusterRouter` functions away from the default network controller, to a different struct / pkg so it can be used by the controllers that require that particular topology (default net and secondary L3 nets). Co-authored-by: Miguel Duarte Barroso <[email protected]> Signed-off-by: Dumitru Ceara <[email protected]>
Signed-off-by: Miguel Duarte Barroso <[email protected]>
Signed-off-by: Miguel Duarte Barroso <[email protected]>
Signed-off-by: Miguel Duarte Barroso <[email protected]>
To introduce minimal changes to the existing unit tests, we keep the default network as was, just renaming its sync method. The default network sync method will in turn invoke the common gatewayManager sync function, which will do all the heavy work. Co-authored-by: Dumitru Ceara <[email protected]> Signed-off-by: Miguel Duarte Barroso <[email protected]>
Signed-off-by: Miguel Duarte Barroso <[email protected]>
This commit is just code plumbing; the actual gw sync function is not being invoked, since we still miss the way to gather the required inputs. Follow-up commits will add these. Co-authored-by: Dumitru Ceara <[email protected]> Signed-off-by: Miguel Duarte Barroso <[email protected]>
Signed-off-by: Dumitru Ceara <[email protected]>
It will be useful in the case of multiple networks that support egress. Signed-off-by: Dumitru Ceara <[email protected]>
This commit builds the GW configuration from multiple sources: - NAD - node annotations The masquerade IPs are generated from the network ID available on the nodes, which is unique for each network, thus guaranteeing that each network also has unique masquerade IPs for it. Co-authored-by: Enrique Llorente <[email protected]> Co-authored-by: Miguel Duarte Barroso <[email protected]> Signed-off-by: Dumitru Ceara <[email protected]>
hack/lint.sh: Mount local golangci-lint cache
This commit refactors the existing code, moving the join switch / cluster router creation away from the default network controller. This is done because the layer2 controllers have no need for this type of topology, and were able to to do previously. This will also make it simpler to unit test the join switch / cluster router creation. We ensure these logical entities have their respective network name in the external IDs, so we a network controller can filter entities for the network it manages. The provided struct is properly unit tested. Signed-off-by: Miguel Duarte Barroso <[email protected]>
We need to use the NAD `networkName` attribute, which will trigger this particular traffic to be sent via a dedicated patch port. Signed-off-by: Miguel Duarte Barroso <[email protected]>
Signed-off-by: Martin Kennelly <[email protected]>
Signed-off-by: Martin Kennelly <[email protected]>
|
@tssurya: This pull request references Jira Issue OCPBUGS-38653, which is valid. The bug has been moved to the POST state. 7 validation(s) were run on this bug
Requesting review from QA contact: This pull request references Jira Issue OCPBUGS-38267, 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 ([email protected]), skipping review request. This pull request references Jira Issue OCPBUGS-38693, which is valid. 7 validation(s) were run on this bug
Requesting review from QA contact: In 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. |
|
/lgtm |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: maiqueb, tssurya The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
1 similar comment
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: maiqueb, tssurya The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
We have https://pr-payload-tests.ci.openshift.org/runs/ci/4219ce90-5eaf-11ef-884f-6a25e6f62e3b-0 and https://pr-payload-tests.ci.openshift.org/runs/ci/4219ce90-5eaf-11ef-884f-6a25e6f62e3b-1 results from payload they look good except for the stable upgrades:
all 3 are showing errors that say: I am not sure what that means in conclusive results? |
|
/label backport-risk-assessed |
openshift-ci
bot
added
the
backport-risk-assessed
openshift-ci
bot
assigned anuragthehatter, asood-rh, huiran0826, jechen0648, mffiedler, qiowang721, rbbratta, weliang1 and yingwang-0320
|
/ocpbugs cc-qa |
|
/label qe-approved |
|
/label cherry-pick-approved |
openshift-ci
bot
added
the
cherry-pick-approved
openshift-merge-bot
bot
merged commit 8c1e320
into
openshift:release-4.17
|
@tssurya: Jira Issue OCPBUGS-38653: All pull requests linked via external trackers have merged: Jira Issue OCPBUGS-38653 has been moved to the MODIFIED state. Jira Issue OCPBUGS-38267: All pull requests linked via external trackers have merged: Jira Issue OCPBUGS-38267 has been moved to the MODIFIED state. Jira Issue OCPBUGS-38693: All pull requests linked via external trackers have merged: Jira Issue OCPBUGS-38693 has been moved to the MODIFIED state. In 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. |
|
[ART PR BUILD NOTIFIER] Distgit: ovn-kubernetes-base |
|
[ART PR BUILD NOTIFIER] Distgit: ovn-kubernetes-microshift |
|
[ART PR BUILD NOTIFIER] Distgit: ose-ovn-kubernetes |
tssurya
changed the title
|
@tssurya: Jira Issue OCPBUGS-38653: All pull requests linked via external trackers have merged: Jira Issue OCPBUGS-38653 has been moved to the MODIFIED state. Jira Issue OCPBUGS-38267 is in an unrecognized state (Verified) and will not be moved to the MODIFIED state. Jira Issue OCPBUGS-38693 is in an unrecognized state (ON_QA) and will not be moved to the MODIFIED state. In 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. |
depends on #2259