Upgrade to Operator SDK 1.41.1 (#211)

* Add missing improvements from Operator SDK 1.34.1

These were not mentioned in the upgrade documentation for version 1.34.x (https://sdk.operatorframework.io/docs/upgrading-sdk-version/v1.34.0/), but I've found them by compating the release with the previous one (https://github.com/operator-framework/operator-sdk/compare/v1.33.0...v1.34.1).

* Upgrade to Operator SDK 1.36.0

Source of upgrade steps: https://sdk.operatorframework.io/docs/upgrading-sdk-version/v1.36.0/
Key differences:
- Go packages `k8s.io/*` are already at a version higher than the one in the upgrade.
- `ENVTEST_K8S_VERSION` is at a version higher than the one in the upgrade
- We didn't have the golangci-lint make command before, thus we only needed to add things.

* Upgrade to Operator SDK 1.38.0

Source of upgrade steps: https://sdk.operatorframework.io/docs/upgrading-sdk-version/v1.38.0/

* Upgrade to Operator SDK 1.39.0

Source of upgrade steps: https://sdk.operatorframework.io/docs/upgrading-sdk-version/v1.39.0/

* Upgrade to Operator SDK 1.40.0

Source of upgrade steps: https://sdk.operatorframework.io/docs/upgrading-sdk-version/v1.40.0/

I didn't do the "Add app.kubernetes.io/name label to your manifests" since it seems that we have it already, and it's customized.

* Address lint errors

* Update golangci-lint version used to support Go 1.24

* Improve workflows

- Make workflow targets more specific.
- Make build workflow only build (i.e. remove test part of it).
- Rearrange steps and improve naming for build workflow.

* Add back deleted test

Initially the test has been removed due to lint saying that it was duplicate code, but it falsely errored since the values are different.

* Improve code and add missing upgrade pieces

* Upgrade to Operator SDK 1.41.1

Source of upgrade steps: https://sdk.operatorframework.io/docs/upgrading-sdk-version/v1.41.0/

Upgrading to 1.41.1 from 1.40.0 doesn't have any migration steps.

Key elements:
- Upgrade to golangci-lint v2
- Made the manifests using the updated controller tools

* Address linter errors

golanci-lint v2 seems to be more robust than the previous one, which is beneficial. Thus, we address the linter errors thrown by v2 and improve our code even further.

* Add Makefile improvements

These were brought in by comparing the Makefile of a freshly created operator using the latest operator-sdk with ours.

* Add missing default kustomization for 1.40.0 upgrade

* Bring default kustomization to latest version

This is done by putting the file's content from a newly-generated operator.

* Switch metrics-bind-address default value back to 8080

This ensures that the upgrade is backwards-compatible.

* Add webhook-related scaffolding

This enables us to easily add support for webhooks by running `operator-sdk create webhook` whenever we want to add them.

* Fix typo
This commit is contained in:
Eduard Filip
2025-07-14 19:32:30 +02:00
committed by GitHub
parent 54eed0c81c
commit cabc020cc6
64 changed files with 1317 additions and 686 deletions

View File

@@ -32,11 +32,19 @@ func SetupConnect(ctx context.Context, kubeClient client.Client, deploymentNames
return nil
}
func setupDeployment(ctx context.Context, kubeClient client.Client, deploymentPath string, deploymentNamespace string) error {
func setupDeployment(
ctx context.Context,
kubeClient client.Client,
deploymentPath string,
deploymentNamespace string,
) error {
existingDeployment := &appsv1.Deployment{}
// check if deployment has already been created
err := kubeClient.Get(ctx, types.NamespacedName{Name: "onepassword-connect", Namespace: deploymentNamespace}, existingDeployment)
err := kubeClient.Get(ctx, types.NamespacedName{
Name: "onepassword-connect",
Namespace: deploymentNamespace,
}, existingDeployment)
if err != nil {
if errors.IsNotFound(err) {
logConnectSetup.Info("No existing Connect deployment found. Creating Deployment")
@@ -46,7 +54,12 @@ func setupDeployment(ctx context.Context, kubeClient client.Client, deploymentPa
return err
}
func createDeployment(ctx context.Context, kubeClient client.Client, deploymentPath string, deploymentNamespace string) error {
func createDeployment(
ctx context.Context,
kubeClient client.Client,
deploymentPath string,
deploymentNamespace string,
) error {
deployment, err := getDeploymentToCreate(deploymentPath, deploymentNamespace)
if err != nil {
return err
@@ -81,8 +94,11 @@ func getDeploymentToCreate(deploymentPath string, deploymentNamespace string) (*
func setupService(ctx context.Context, kubeClient client.Client, servicePath string, deploymentNamespace string) error {
existingService := &corev1.Service{}
//check if service has already been created
err := kubeClient.Get(ctx, types.NamespacedName{Name: "onepassword-connect", Namespace: deploymentNamespace}, existingService)
// check if service has already been created
err := kubeClient.Get(ctx, types.NamespacedName{
Name: "onepassword-connect",
Namespace: deploymentNamespace,
}, existingService)
if err != nil {
if errors.IsNotFound(err) {
logConnectSetup.Info("No existing Connect service found. Creating Service")
@@ -92,7 +108,12 @@ func setupService(ctx context.Context, kubeClient client.Client, servicePath str
return err
}
func createService(ctx context.Context, kubeClient client.Client, servicePath string, deploymentNamespace string) error {
func createService(
ctx context.Context,
kubeClient client.Client,
servicePath string,
deploymentNamespace string,
) error {
f, err := os.Open(servicePath)
if err != nil {
return err