Update perfectscale-agent chart with new package version v1.0.79

Author: github-actions

charts/perfectscale-agent/Chart.yaml

@@ -15,4 +15,4 @@ dependencies:
     repository: https://prometheus-community.github.io/helm-charts
     condition: settings.deployWindowsExporter
 annotations:
-  charts.openshift.io/name: exporter
+  charts.openshift.io/name: exporter

charts/perfectscale-agent/crds/clustersettings.yaml

@@ -0,0 +1,285 @@
+# ClusterSettings CRD Examples
+
+This directory contains examples of how to use the `ClusterSettings` Custom Resource Definition (CRD) to configure PerfectScale profiles via GitOps.
+
+## Recommended Workflow
+
+**We prefer users to use refByID rather than inline profiles.**
+
+1. **Create and test profiles in UI first**
+   - Create profiles through the PerfectScale UI
+   - Test and validate the configuration
+   - Once satisfied, note the profile ID shown in UI (format: {type|class}-N, e.g., jira-1, slack-2)
+
+2. **Reference tested profiles via refByID in CR (preferred method)**
+   - Use `refByID` to reference the UI-created profile by its ID
+   - No `assigned` field needed - refByID profiles are always assigned
+   - See examples in `clustersettings-assign-existing.yaml`
+
+3. **Alternative: Create inline profiles for testing**
+   - Create inline profiles with `assigned: false` (default)
+   - Profiles will be visible in UI and can be tested
+   - Once tested, update CR to set `assigned: true` to activate the profile
+
+## Validation Rules
+
+- **Only ONE profile of each type can be assigned to a cluster**
+- **For inline profiles**: Only ONE can have `assigned: true` per type (validation fails if multiple)
+- **For refByID profiles**: Only ONE can be specified per type (validation fails if multiple)
+- **IMPORTANT**: You CANNOT have BOTH a refByID AND an inline profile with `assigned: true` for the same type
+  - If refByID exists for a type, ALL inline profiles of that type MUST have `assigned: false` (or unset)
+  - Having both will fail validation
+  - Valid: refByID + inline profiles with `assigned: false`
+  - Invalid: refByID + any inline profile with `assigned: true`
+- **refByID profiles**: Always considered assigned (no `assigned` field needed or required)
+- **Non-existent refByID**: If referenced profile doesn't exist, validation passes but nothing happens (noop)
+- **CRD Override**: CRD profiles (both refByID and inline) ALWAYS WIN over UI-assigned profiles
+
+## Important Notes
+
+- **All ClusterSettings resources MUST be applied in the `perfectscale` namespace**
+- Only **one** ClusterSettings resource is allowed per cluster
+- Secrets referenced in the CRD must exist in the `perfectscale` namespace
+- **Profiles assigned via CRD cannot be deleted via UI** - deletion must be done via CRD
+
+## UI-only Features
+
+Some PerfectScale features are currently available only through the UI and cannot be configured via CRD:
+- **Custom Grouping** (Ephemeral Pods Grouping) - See: https://docs.perfectscale.io/customize-workflow/ephemeral-pods-grouping
+- **Automation Policies** - Automation rules and maintenance windows
+
+All other profiles are fully supported via CRD.
+
+## Files Overview
+
+### Complete Examples
+- **`clustersettings-all.yaml`** - Comprehensive example showing all profile types and configuration options
+- **`clustersettings-mixed.yaml`** - Example combining inline profiles and profile references
+
+### Simple Examples
+- **`clustersettings-assign-existing.yaml`** - Assign existing profiles to cluster
+- **`clustersettings-custom-pricing.yaml`** - Create custom node pricing inline
+- **`clustersettings-podfit-labels.yaml`** - Configure PodFit label visibility
+- **`clustersettings-customization.yaml`** - Configure workload labels
+- **`clustersettings-resiliency-alerts.yaml`** - Resiliency alerts configuration
+
+### Pricing Examples
+- **`clustersettings-aws-cur.yaml`** - AWS Cost and Usage Report integration
+- **`clustersettings-gcp-billing.yaml`** - GCP billing export integration
+- **`clustersettings-azure-billing.yaml`** - Azure billing integration
+- **`clustersettings-cloud-billing.yaml`** - Multi-cloud billing (GCP + Azure)
+
+### Integration Examples
+- **`clustersettings-slack.yaml`** - Slack integration for alerts
+- **`clustersettings-teams.yaml`** - Microsoft Teams integration
+- **`clustersettings-jira.yaml`** - Jira ticketing integration
+- **`clustersettings-datadog.yaml`** - Datadog monitoring integration
+- **`clustersettings-observability.yaml`** - Custom observability platform
+- **`clustersettings-integrations.yaml`** - Multiple integrations (Teams + Jira + Datadog)
+
+## Profile Types
+
+### Pricing Profiles
+Control how costs are calculated for your cluster.
+
+Optional field for any pricing profile:
+- `global_discount` - Applies a single discount percent to on-demand resource prices for any pricing profile type (not applied to DoiT pricing profiles). Discount applies only when `start_date` is provided and valid.
+  - `percentage` - Discount percent (0-100) applied to on-demand resource prices only.
+  - `start_date` - ISO date (yyyy-mm-dd) when the discount starts; if missing or invalid, the discount is not applied. Changing this date does not automatically update past reports; contact support if historical data needs recalculation.
+
+**Types:**
+- `custom` - Define custom pricing per instance type
+  - `instanceType`: The name of the instance that is given by the provider (e.g., c5.large, m5.xlarge)
+  - `memGBHourPrice`: Indicates the price ($) of 1 GB of memory per hour
+  - `cpuCoreHourPrice`: Indicates the price ($) of 1 core per hour
+  - `nodeHourPrice`: Indicates the price ($) of a node per hour
+- `aws_cur` - AWS Cost and Usage Report integration
+  - Authentication Method 1 (Recommended): IAM Role
+    - `role_arn`: The Amazon Resource Name (ARN) associated with the role possessing the necessary credentials to execute calls on your behalf
+    - `aws_external_id`: The ID for cross-account access in AWS Identity and Access Management (IAM). A unique, user-defined string that ensures only trusted third-party entities can assume a specific role
+  - Authentication Method 2 (Alternative): IAM User
+    - `access_key_id`: AWS access key ID (via `access_key_id` in CRD)
+    - `secret_access_key`: AWS secret access key (via `secret_access_key_from` secret reference in CRD)
+  - Configuration fields:
+    - `athena_database`: Name of the database that was created during the Athena setup
+    - `athena_region`: AWS region where Athena is running
+    - `athena_result_bucket`: S3 bucket where Athena stores query results
+    - `athena_table`: Name of the table that was created on the Athena setup
+    - `aws_account_id`: AWS account where cluster is running (NOT the AWS billing account ID)
+  - Full guide: https://docs.perfectscale.io/cloud-billing-integration/connecting-aws-cur
+- `gcp_billing` - GCP BigQuery billing export integration
+- `azure_billing` - Azure billing integration
+
+### Integration Profiles
+Configure external systems for notifications and monitoring.
+
+**Types:**
+- `slack` - Slack notifications
+  - `slack_token`: A mandatory field used to permit PerfectScale to interact with your Slack (via `slack_token_from` secret reference in CRD)
+    - Setup: Create an app at https://api.slack.com/apps
+    - Required scopes: chat:write.public, chat:write, channels:read, conversations.list
+  - `channel`: Slack channel name that indicates where to receive PerfectScale alerts
+  - `routings`: Optional field allows alerts to be sent to different Slack channels for different workloads in a cluster
+- `teams` - Microsoft Teams notifications
+  - `webhook_url`: A mandatory field used to permit PerfectScale to interact with your Teams (via `webhook_url_from` secret reference in CRD)
+    - Setup: Go to Apps at https://teams.microsoft.com/, add Incoming Webhook, and connect to a specific channel
+  - `routings`: Optional field allows alerts to be sent for specific alert types
+- `jira` - Jira ticketing
+- `datadog` - Datadog monitoring
+- `observability` - Custom observability platform
+  - Setup: Go to https://github.com/perfectscale-io/observability, download either the Grafana or DataDog dashboard, import into your instances, and copy the dashboard link
+
+### Resiliency Alerts Profiles
+Configure workload health and resiliency alerts. **No type field required.**
+
+### Podfit Labels Profiles
+Configure which labels are displayed in the PodFit view. **No type field required.**
+
+### Customization Profiles
+Configure custom workload labels. **No type field required.**
+
+**Fields:**
+- `workload_labels`: Defines labels that can be applied to Kubernetes resources (pods, deployments, or stateful sets) to identify and manage workloads within a cluster
+- `node.spot_labels`: Identifies spot nodes (key-value pairs)
+- `node.node_group_labels`: Identifies the nodes group in a cluster
+- `node.architecture_labels`: Specifies the architecture of the node (x86_64, arm64, etc.)
+- `node.os_labels`: Specifies the operating system running on a node within the Kubernetes cluster
+- `node.instance_type_labels`: Specifies the type or configuration of VMs (instances) within a cloud infrastructure
+- `node.region_labels`: Specifies the geographical region where a particular resource is located
+- `node.availability_zone_labels`: Specify the availability zone where a particular resource is located
+- `ignored_pod_labels`: Defines labels that Kubernetes components should disregard when making pod decisions
+- `ignored_node_labels`: Defines labels that Kubernetes components should disregard when making node decisions
+
+**NOTE:** Labels should follow Prometheus conventions. Learn more at: https://prometheus.io/docs/concepts/data_model/#metric-names-and-labels
+
+## Profile Assignment Methods
+
+### Method 1: Reference Existing Profile (refByID)
+Reference a profile that was created in the UI:
+
+```yaml
+profiles:
+  pricing:
+    - type: custom
+      refByID: "custom-1"  # ID format: {type|class}-N
+```
+
+**Note:** Profiles created via CRD are cluster-specific and cannot be referenced by other clusters.
+
+**Important Rules:**
+- `refByID` can only reference UI-created profiles (not inline CRD profiles)
+- Profile IDs are shown in UI in format {type|class}-N (e.g., jira-1, slack-2, custom-3)
+- **IMPORTANT**: Underscores in type/class names are converted to hyphens in IDs:
+  - Class `resiliency_alerts` → ID `resiliency-alerts-1`
+  - Class `podfit_labels` → ID `podfit-labels-1`
+  - Type `aws_cur` → ID `aws-cur-1`
+  - Type `gcp_billing` → ID `gcp-billing-1`
+  - Type `azure_billing` → ID `azure-billing-1`
+- The type field must match the referenced profile's type
+- refByID profiles are always assigned (no `assigned` field needed)
+- If the referenced profile doesn't exist, the assignment won't happen
+- If the referenced profile is updated in UI, changes propagate to referencing clusters
+- Only ONE profile per class can have `assigned: true` (applies to inline profiles only)
+
+### Method 2: Create Inline Profile
+Create a new profile directly in the CRD:
+
+```yaml
+profiles:
+  pricing:
+    - type: custom
+      name: "my-inline-profile"
+      assigned: true
+      value:
+        nodeTypes:
+          - instanceType: c5.large
+            pricing:
+              cpuCoreHourPrice: 0.08
+              memGBHourPrice: 0.02
+```
+
+**Note:** Inline profiles are cluster-specific and only available to the cluster where they are created.
+
+## Using Secrets
+
+You can't specify sensitive data directly in the CRD. Use secret references instead.
+
+```yaml
+secret_access_key_from:
+  secretKeyRef:
+    name: aws-credentials  # Secret name in perfectscale namespace
+    key: secret-access-key # Key within the secret
+```
+
+## Validation Rules
+
+1. **Singleton:** Only one ClusterSettings resource per cluster
+2. **Valid Types:** Profile types must be valid for their class
+3. **Name Requirements:**
+   - `refByID` cannot be empty if provided
+   - Inline profiles must have a `name`
+4. **Assignment:** Only one inline profile per class can have `assigned: true`
+5. **Type Requirements:**
+   - `pricing` and `integrations` MUST have a `type`
+   - `resiliency_alerts`, `podfit_labels`, `customization` MUST NOT have a `type`
+
+## Applying Examples
+
+```bash
+# Apply all resources (including secrets) in the example file
+kubectl apply -f examples/clustersettings-aws-cur.yaml
+
+# View the ClusterSettings
+kubectl get clustersettings -n perfectscale
+
+# Describe to see details
+kubectl describe clustersettings cluster-settings-main -n perfectscale
+
+# Check logs for processing status
+kubectl logs -n perfectscale -l app=perfectscale-exporter --tail=100
+```
+
+## Deleting Configuration
+
+When you delete the ClusterSettings resource, the cluster returns to having no profiles (or default profiles):
+
+```bash
+kubectl delete clustersettings cluster-settings-main -n perfectscale
+```
+
+## Documentation Links
+
+All profile types are documented in detail:
+
+### Pricing
+- **Custom Pricing**: https://docs.perfectscale.io/customize-workflow/pricing/custom-pricing-configuration
+- **AWS CUR**: https://docs.perfectscale.io/cloud-billing-integration/connecting-aws-cur
+- **GCP Billing**: https://docs.perfectscale.io/customize-workflow/pricing/gcp-cloud-billing-configuration
+- **Azure Billing**: https://docs.perfectscale.io/cloud-billing-integration/connecting-azure-cost-management
+
+### Integrations
+- **Slack**: https://docs.perfectscale.io/customize-workflow/communication-and-messaging/slack-integration
+- **Teams**: https://docs.perfectscale.io/customize-workflow/communication-and-messaging/ms-teams-integration
+- **Jira**: https://docs.perfectscale.io/customizations/ticketing-and-bug-tracking
+- **Datadog**: https://docs.perfectscale.io/customize-workflow/communication-and-messaging/datadog-alerts-integration
+- **Observability**: https://docs.perfectscale.io/customize-workflow/observability
+
+### Customizations
+- **Label Customizations**: https://docs.perfectscale.io/customizations/label-customizations
+- **PodFit Labels**: https://docs.perfectscale.io/customize-workflow/podfit-labels
+- **Resiliency Alerts**: https://docs.perfectscale.io/customize-workflow/alerting/resiliency-alerts
+
+## Troubleshooting
+
+### CRD Creation Failed
+- Check if ClusterSettings already exists: `kubectl get clustersettings -n perfectscale`
+- Review logs: `kubectl logs -n perfectscale -l app=perfectscale-exporter`
+
+### Profile Not Applied
+- Check logs for sync errors
+- Verify secrets exist: `kubectl get secrets -n perfectscale`
+- Check profile format matches the type
+
+### Secret Not Found
+- Verify secret exists: `kubectl get secret <secret-name> -n perfectscale`
+- Ensure secret is in `perfectscale` namespace