Update dependency rancher/cluster-api-provider-rke2 to v0.12.0 + custom patch (main) (!3640) · Merge requests · Sylva-projects / sylva-core

Only the base of this MR was generated by Renovate Bot Sylva instance 🤖. The rest was by @tmmorin.

Closes #2060 (closed)

With this MR, we still do not use an official/vanilla version of RKE2 CAPI provider. See details in #1931 (closed).

What was done additionally to upgrading the version:

prepare a custom build of the CAPI RKE2 controlplane provider to skip the removal of etcd membership on a node being deleted
update where infrastructureRef is set in RKE2ControlPlane
- (details explained in sylva-projects/sylva-elements/helm-charts/sylva-capi-cluster!620 (merged))
workaround for https://github.com/rancher/cluster-api-provider-rke2/issues/596
- When upgrading RKE2 CAPI provider from 0.7.1 to 0.12.0, the newer controller instantly starts a node rolling update on all clusters (mgmt cluster and all workload clusters).
- We don't want this:
  - for the mgmt cluster, no node rolling update should happen before we modify CAPI resources in sylva-capi-cluster
  - for workload clusters, similarly, we don't want any node rolling update to be triggered just because we changed something in the mgmt cluster
- To avoid this, our workaround is to:
  - before upgrading cabpr unit, we set spec.paused: true on the Cluster resources of all clusters with a new cabpr-cluster-pause kube-job unit (introduced by !3640 (merged))
  - when cabpr is upgraded, since all clusters are paused, nothing happens
  - to un-pause a cluster after its definition has been updated, we add a post-upgrade hook removing spec.paused (done in sylva-projects/sylva-elements/helm-charts/sylva-capi-cluster!620 (merged))

This MR depends on sylva-projects/sylva-elements/helm-charts/sylva-capi-cluster!620 (merged)

This MR contains the following updates:

Package	Update	Change
rancher/cluster-api-provider-rke2	minor	`v0.7.1` -> `v0.12.0`

Release Notes

CI configuration

Below you can choose test deployment variants to run in this MR's CI.

Click to open to CI configuration

Legend:

Icon	Meaning	Available values
☁️	Infra Provider	`capd`, `capo`, `capm3`
🚀	Bootstrap Provider	`kubeadm` (alias `kadm`), `rke2`
🐧	Node OS	`ubuntu`, `suse`
🛠️	Deployment Options	`light-deploy`, `dev-sources`, `ha`, `misc`
🎬	Pipeline Scenarios	Available scenario list and description

Global config for deployment pipelines

autorun pipelines
allow failure on pipelines
record sylvactl events

Notes:

Enabling autorun will make deployment pipelines to be run automatically without human interaction
Disabling allow failure will make deployment pipelines mandatory for pipeline success.
if both autorun and allow failure are disabled, deployment pipelines will need manual triggering but will be blocking the pipeline

Be aware: after configuration change, pipeline is not triggered automatically. Please run it manually (by clicking the run pipeline button in Pipelines tab) or push new code.

⚙️ If no checkbox is checked, a default pipeline will be enabled (capm3, or capo if capo label is set)

Edited Mar 18, 2025 by Thomas Morin

Update dependency rancher/cluster-api-provider-rke2 to v0.12.0 + custom patch (main)

Release Notes

CI configuration

Global config for deployment pipelines

Merge request reports