Telemetry enhancement

[minjieqiu]

Description.

This PR implement SOK Telemetry enhancement. ERD:
https://cisco-my.sharepoint.com/:w:/p/mqiu/IQBoVUuEEY1SR4rDjbja0iPuAeN5dxFG-K-ZPpvO6RoWJp0?e=n5R1Ow

What does this PR have in it?.

Periodically collect (once per day) and send SOK telemetry which includes:

1. SOK telemetry.
   a. SOK version.
   b. CPU/Memory settings (limit and request) of containers including standalone, searchheadcluster, indexercluster,
   clustermaster, clustermanager, licensemaster and licensemanager.
   c. LincenseInfo (Splunk license ID and license type).
2. Other component's telemetry which are submitted to SOK by adding key/value to the new telemetry configmap splunk-operator-manager-telemetry

Key Changes.

• Created a new configmap splunk-operator-manager-telemetry
• Create a new controller which reconciles on the telemetry configmap
• Renamed the telemetry app to app_tel_for_sok

Highlight the updates in specific files

Testing and Verification.

Tested on s1, c3 and m4.

How did you test these changes? What automated tests are added?.
Added telemetry verification to existing s1, c3 and m4 tests.

Related Issues

Jira tickets, GitHub issues, Support tickets...
https://splunk.atlassian.net/browse/CSPL-4371.

PR Checklist

• [✅ ] Code changes adhere to the project's coding standards.
• [ ✅ ] Relevant unit and integration tests are included.
• [✅ ] Documentation has been updated accordingly.
• [✅ ] All tests pass locally.
• [✅ ] The PR description follows the project's guidelines.

[github-actions]

CLA Assistant Lite bot All contributors have signed the CLA ✍️ ✅

[minjieqiu]

I have read the CLA Document and I hereby sign the CLA

[minjieqiu]

I have read the Code of Conduct and I hereby accept the Terms

[coveralls]

Pull Request Test Coverage Report for Build 21975003576

Details

• 372 of 465 (80.0%) changed or added relevant lines in 6 files are covered.
• 3 unchanged lines in 1 file lost coverage.
• Overall coverage decreased (-0.3%) to 86.022%

---

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
pkg/splunk/client/enterprise.go	13	15	86.67%
pkg/splunk/enterprise/names.go	0	6	0.0%
internal/controller/telemetry_controller.go	37	46	80.43%
pkg/splunk/enterprise/telemetry.go	316	392	80.61%

Files with Coverage Reduction	New Missed Lines	%
pkg/splunk/enterprise/afwscheduler.go	3	92.51%

Totals
Change from base Build 21948077040:	-0.3%
Covered Lines:	11293
Relevant Lines:	13128

---

💛 - Coveralls

[kasiakoziol]

I think it might be worth to add/update docs

[vivekr-splunk]

we should have some controller test cases

[vivekr-splunk]

should we refactor this code to make it much easier to read, or use generics
an example

func collectDeploymentTelDataRefactored(ctx context.Context, client splcommon.ControllerClient, deploymentData map[string]interface{}) map[string][]splcommon.MetaObject {
	reqLogger := log.FromContext(ctx)
	scopedLog := reqLogger.WithName("collectDeploymentTelData")

	crWithTelAppList := make(map[string][]splcommon.MetaObject)
	scopedLog.Info("Start collecting deployment telemetry data")

	// Define all CR handlers in a slice
	handlers := []crListHandler{
		{kind: "Standalone", listFunc: listStandalones, checkTelApp: true},
		{kind: "LicenseManager", listFunc: listLicenseManagers, checkTelApp: true},
		{kind: "LicenseMaster", listFunc: listLicenseMasters, checkTelApp: true},
		{kind: "SearchHeadCluster", listFunc: listSearchHeadClusters, checkTelApp: true},
		{kind: "IndexerCluster", listFunc: listIndexerClusters, checkTelApp: false},
		{kind: "ClusterManager", listFunc: listClusterManagers, checkTelApp: true},
		{kind: "ClusterMaster", listFunc: listClusterMasters, checkTelApp: true},
		{kind: "MonitoringConsole", listFunc: listMonitoringConsoles, checkTelApp: false},
	}

	// Process each CR type using the same logic
	for _, handler := range handlers {
		processCRType(ctx, client, handler, deploymentData, crWithTelAppList, scopedLog)
	}

	return crWithTelAppList
}

// processCRType is the common processing logic for all CR types
func processCRType(
	ctx context.Context,
	client splcommon.ControllerClient,
	handler crListHandler,
	deploymentData map[string]interface{},
	crWithTelAppList map[string][]splcommon.MetaObject,
	scopedLog interface{}, // Using interface{} to avoid import issues, should be logr.Logger
) {
	items, err := handler.listFunc(ctx, client)
	if err != nil {
		// scopedLog.Error(err, "Failed to list objects", "kind", handler.kind)
		return
	}

	if len(items) == 0 {
		return
	}

	// Create per-kind data map
	perKindData := make(map[string]interface{})
	deploymentData[handler.kind] = perKindData

	// Process each item
	for _, item := range items {
		// scopedLog.Info("Collecting data", "kind", item.kind, "name", item.name, "namespace", item.namespace)

		crResourceData := make(map[string]string)
		perKindData[item.name] = crResourceData

		// Collect resource telemetry data
		if resources, ok := item.resources.(corev1.ResourceRequirements); ok {
			collectResourceTelData(resources, crResourceData)
		}

		// Add to telemetry app list if applicable
		if handler.checkTelApp && item.hasTelApp {
			crWithTelAppList[handler.kind] = append(crWithTelAppList[handler.kind], item.cr)
		} else if handler.checkTelApp && !item.hasTelApp {
			// scopedLog.Info("Telemetry app is not installed for this CR", "kind", item.kind, "name", item.name)
		}
	}
}

// List functions for each CR type - these extract the common pattern

func listStandalones(ctx context.Context, client splcommon.ControllerClient) ([]crItem, error) {
	var list enterpriseApi.StandaloneList
	err := client.List(ctx, &list)
	if err != nil {
		return nil, err
	}

	items := make([]crItem, 0, len(list.Items))
	for i := range list.Items {
		cr := &list.Items[i]
		items = append(items, crItem{
			name:      cr.GetName(),
			namespace: cr.GetNamespace(),
			kind:      cr.Kind,
			resources: cr.Spec.CommonSplunkSpec.Resources,
			hasTelApp: cr.Status.TelAppInstalled,
			cr:        cr,
		})
	}
	return items, nil
}

func listLicenseManagers(ctx context.Context, client splcommon.ControllerClient) ([]crItem, error) {
	var list enterpriseApi.LicenseManagerList
	err := client.List(ctx, &list)
	if err != nil {
		return nil, err
	}

	items := make([]crItem, 0, len(list.Items))
	for i := range list.Items {
		cr := &list.Items[i]
		items = append(items, crItem{
			name:      cr.GetName(),
			namespace: cr.GetNamespace(),
			kind:      cr.Kind,
			resources: cr.Spec.CommonSplunkSpec.Resources,
			hasTelApp: cr.Status.TelAppInstalled,
			cr:        cr,
		})
	}
	return items, nil
}

func listLicenseMasters(ctx context.Context, client splcommon.ControllerClient) ([]crItem, error) {
	var list enterpriseApiV3.LicenseMasterList
	err := client.List(ctx, &list)
	if err != nil {
		return nil, err
	}

	items := make([]crItem, 0, len(list.Items))
	for i := range list.Items {
		cr := &list.Items[i]
		items = append(items, crItem{
			name:      cr.GetName(),
			namespace: cr.GetNamespace(),
			kind:      cr.Kind,
			resources: cr.Spec.CommonSplunkSpec.Resources,
			hasTelApp: cr.Status.TelAppInstalled,
			cr:        cr,
		})
	}
	return items, nil
}

func listSearchHeadClusters(ctx context.Context, client splcommon.ControllerClient) ([]crItem, error) {
	var list enterpriseApi.SearchHeadClusterList
	err := client.List(ctx, &list)
	if err != nil {
		return nil, err
	}

	items := make([]crItem, 0, len(list.Items))
	for i := range list.Items {
		cr := &list.Items[i]
		items = append(items, crItem{
			name:      cr.GetName(),
			namespace: cr.GetNamespace(),
			kind:      cr.Kind,
			resources: cr.Spec.CommonSplunkSpec.Resources,
			hasTelApp: cr.Status.TelAppInstalled,
			cr:        cr,
		})
	}
	return items, nil
}

func listIndexerClusters(ctx context.Context, client splcommon.ControllerClient) ([]crItem, error) {
	var list enterpriseApi.IndexerClusterList
	err := client.List(ctx, &list)
	if err != nil {
		return nil, err
	}

	items := make([]crItem, 0, len(list.Items))
	for i := range list.Items {
		cr := &list.Items[i]
		items = append(items, crItem{
			name:      cr.GetName(),
			namespace: cr.GetNamespace(),
			kind:      cr.Kind,
			resources: cr.Spec.CommonSplunkSpec.Resources,
			hasTelApp: false, // IndexerClusters don't track TelAppInstalled
			cr:        cr,
		})
	}
	return items, nil
}

func listClusterManagers(ctx context.Context, client splcommon.ControllerClient) ([]crItem, error) {
	var list enterpriseApi.ClusterManagerList
	err := client.List(ctx, &list)
	if err != nil {
		return nil, err
	}

	items := make([]crItem, 0, len(list.Items))
	for i := range list.Items {
		cr := &list.Items[i]
		items = append(items, crItem{
			name:      cr.GetName(),
			namespace: cr.GetNamespace(),
			kind:      cr.Kind,
			resources: cr.Spec.CommonSplunkSpec.Resources,
			hasTelApp: cr.Status.TelAppInstalled,
			cr:        cr,
		})
	}
	return items, nil
}

func listClusterMasters(ctx context.Context, client splcommon.ControllerClient) ([]crItem, error) {
	var list enterpriseApiV3.ClusterMasterList
	err := client.List(ctx, &list)
	if err != nil {
		return nil, err
	}

	items := make([]crItem, 0, len(list.Items))
	for i := range list.Items {
		cr := &list.Items[i]
		items = append(items, crItem{
			name:      cr.GetName(),
			namespace: cr.GetNamespace(),
			kind:      cr.Kind,
			resources: cr.Spec.CommonSplunkSpec.Resources,
			hasTelApp: cr.Status.TelAppInstalled,
			cr:        cr,
		})
	}
	return items, nil
}

func listMonitoringConsoles(ctx context.Context, client splcommon.ControllerClient) ([]crItem, error) {
	var list enterpriseApi.MonitoringConsoleList
	err := client.List(ctx, &list)
	if err != nil {
		return nil, err
	}

	items := make([]crItem, 0, len(list.Items))
	for i := range list.Items {
		cr := &list.Items[i]
		items = append(items, crItem{
			name:      cr.GetName(),
			namespace: cr.GetNamespace(),
			kind:      cr.Kind,
			resources: cr.Spec.CommonSplunkSpec.Resources,
			hasTelApp: false, // MonitoringConsoles don't track TelAppInstalled
			cr:        cr,
		})
	}
	return items, nil
}

[minjieqiu]

Thanks for the code suggestion. I have made the change.

[vivekr-splunk]

code has 47% test coverage lets try to move to 90%

[minjieqiu]

I have added more tests.

[kubabuczak]

Is this hardcoded?

[minjieqiu]

Yes. This config map is not accessed by multiple CRs.