🐛 Enhance OpenStackMachine controller to handle cases where ControlPlane is not managed by CAPO (user provided ports, SGs etc) #2381
Conversation
Signed-off-by: Bharath Nallapeta <[email protected]>
k8s-ci-robot
added
the
do-not-merge/hold
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
k8s-ci-robot
added
the
cncf-cla: yes
|
Welcome @bnallapeta! |
k8s-ci-robot
added
the
needs-ok-to-test
|
Hi @bnallapeta. Thanks for your PR. I'm waiting for a kubernetes-sigs member to verify that this patch is reasonable to test. If it is, they should reply with Once the patch is verified, the new status will be reflected by the I understand the commands that are listed here. 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-sigs/prow repository. |
k8s-ci-robot
added
the
size/L
✅ Deploy Preview for kubernetes-sigs-cluster-api-openstack ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
|
/ok-to-test |
k8s-ci-robot
added
ok-to-test
There was a problem hiding this comment.
I'm in favour of this in principal. I know @EmilienM was bitten by issues with Network not being set in a bunch of places, though. We also need to consider:
- The bastion host
- All possible variations of the delete flow (i.e. delete before creation, partial creation, after creation)
How can we add this to our E2E tests? This might be tricky because I don't think we can currently create an OpenStackCluster which doesn't create a network. However, if it's not in the E2E tests I would assume that it doesn't work.
| defaultNetworkID := "" | ||
| if len(openStackMachine.Spec.Ports) == 0 { | ||
| if openStackCluster.Status.Network == nil || openStackCluster.Status.Network.ID == "" { | ||
| return nil, fmt.Errorf( |
There was a problem hiding this comment.
Please can you wrap this in a TerminalError so we'll stop trying to reconcile it?
There was a problem hiding this comment.
I believe the OpenStackServer controller has a great example.
|
@bnallapeta: The following test failed, say
Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR. 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-sigs/prow repository. I understand the commands that are listed here. |
| machineServerSpec := openStackMachineSpecToOpenStackServerSpec(&openStackMachine.Spec, identityRef, compute.InstanceTags(&openStackMachine.Spec, openStackCluster), failureDomain, userDataRef, getManagedSecurityGroup(openStackCluster, machine), openStackCluster.Status.Network.ID) | ||
| // If user has provided .spec.ports (non-empty), skip .Status.Network entirely | ||
| // Otherwise, fall back to openStackCluster.Status.Network.ID | ||
| // This supports HCP - Hosted Control Plane usage |
There was a problem hiding this comment.
Let's write a more generic comment, please.
In general, let's remove "HPC / HCP" from the whole PR. Let's consider it a generic use case where the control plane is not managed by CAPO.
bnallapeta
changed the title
What this PR does / why we need it:
Hosted Control Plane (HCP) clusters do not use a control-plane node in OpenStack, so CAPO should rely on user-specified ports and security groups rather than cluster defaults.
Which issue(s) this PR fixes (optional, in
fixes #<issue number>(, fixes #<issue_number>, ...)format, will close the issue(s) when PR gets merged):Fixes #2380
Special notes for your reviewer:
Avoids nil-pointer panics for HPC setups and lets users fully customize network and security groups for worker nodes.
TODOs:
/hold