NP-764: Enable no uplink OVN-K gateway bridge mode#1692
NP-764: Enable no uplink OVN-K gateway bridge mode#1692openshift-merge-robot merged 2 commits intoopenshift:mainfrom
Conversation
|
@pliurh I think that if you rebase this PR on the HEAD of the main branch we will get better results from the CI jobs. |
3745816 to
181e9fe
Compare
|
/retest |
|
/retest |
1 similar comment
|
/retest |
|
/hold This is failing on QE Ushift clusters causing node to go into NOTREADY state with error
|
|
/retest |
1 similar comment
|
/retest |
|
Today's rebase brought in changes for amd64 (it seems that we already had them for arm - introduced in 2023-06-15-173508, we had |
|
some infra problems |
|
/test e2e-openshift-conformance-reduced |
Yes.
Agree.
The |
0f181ab to
baefef1
Compare
There was a problem hiding this comment.
In a fresh deployment, the service address will always be assigned to loopback device since this node network configuration runs before the InfrastructureServices controller (where CNI is started). Is it expected?
|
/retest |
|
@jcaamano Could you also take a look? |
|
@pliurh: This pull request references NP-764 which is a valid jira issue. 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 kubernetes/test-infra repository. |
|
/unhold |
That part LGTM, just that small comment |
|
/retest |
Use no uplink OVN-K gateway bridge to decouple br-ex from host physical interface. Use ovs-vsctl to create br-ex bridge instead of NM, as we don't need run DHCP client for br-ex interface anymore. Signed-off-by: Peng Liu <pliu@redhat.com>
Signed-off-by: Peng Liu <pliu@redhat.com>
|
/retest |
1 similar comment
|
/retest |
|
/test e2e-openshift-conformance-reduced-arm |
dhellmann
left a comment
There was a problem hiding this comment.
This version looks good and seems to be passing the tests now that the topolvm image issue is resolved. I have one comment about some documentation, but that can (and probably should) be addressed in a simple follow-up PR, rather than requiring another iteration on this one.
The ARM tests are still a bit flakey for reasons outside of our control, so if those fail we can still merge this, unless you have reason to think the failure is related to the code.
@zshi-redhat do you want to apply the LGTM when you're happy?
There was a problem hiding this comment.
The details in this comment thread are an important part of the design. Are they written down somewhere? If not, could you add them in a comment in this function, please? You can use a follow-up PR, since this one looks like it is almost ready to be merged.
@pliurh please file a followup PR for the documentation change.
agree, looking at the job history, it has 4 pass in May and zero pass in June. And the failed tests are different in each run. I will go ahead and lgtm the PR.
/lgtm |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: pliurh, zshi-redhat The full list of commands accepted by this bot can be found here. The pull request process is described here DetailsNeeds approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
@pliurh: The following test 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/test-infra repository. I understand the commands that are listed here. |
Use no uplink OVN-K gateway bridge to decouple br-ex from host physical interface.
Use ovs-vsctl to create br-ex bridge instead of NM, as we don't need to run DHCP client for br-ex interface anymore.