chore: update doc

Author: hccheng72

deploy/example/metrics/README.md

@@ -7,8 +7,8 @@ kubectl apply -f https://raw.githubusercontent.com/kubernetes-sigs/azurefile-csi
 2. Get `EXTERNAL-IP` of service `csi-azurefile-controller`
 ```console
 $ kubectl get svc csi-azurefile-controller -n kube-system
-NAME                       TYPE        CLUSTER-IP   EXTERNAL-IP   PORT(S)     AGE
-csi-azurefile-controller   ClusterIP   10.0.184.0   20.39.21.132  29614/TCP   47m
+NAME                       TYPE           CLUSTER-IP   EXTERNAL-IP    PORT(S)           AGE
+csi-azurefile-controller   LoadBalancer   10.0.184.0   20.39.21.132   29614:30563/TCP   47m
 ```
 
 3. Run following command to get cloudprovider_azure operation metrics
@@ -17,8 +17,58 @@ ip=`kubectl get svc csi-azurefile-controller -n kube-system | grep file | awk '{
 curl http://$ip:29614/metrics | grep cloudprovider_azure | grep file | grep -e sum -e count
 ```
 
+4. Run following command to get CSI-specific operation metrics
+```console
+ip=`kubectl get svc csi-azurefile-controller -n kube-system | grep file | awk '{print $4}'`
+curl http://$ip:29614/metrics | grep azurefile_csi_driver_operation | grep -e sum -e count
+```
+
+
+## CSI Driver Metrics
+
+The Azure File CSI driver exposes the following custom metrics:
+
+### Controller Metrics (port 29614)
+
+| Metric | Type | Labels | Description |
+|--------|------|--------|-------------|
+| `azurefile_csi_driver_operation_duration_seconds` | Histogram | `operation`, `success` | Duration of CSI operations in seconds |
+| `azurefile_csi_driver_operation_duration_seconds_labeled` | Histogram | `operation`, `success`, `protocol`, `storage_account_type` | Duration of CSI operations with additional labels |
+| `azurefile_csi_driver_operations_total` | Counter | `operation`, `success` | Total number of CSI operations |
+
+**Label Values:**
+- `operation`: `controller_create_volume`, `controller_delete_volume`, `controller_create_snapshot`, `controller_delete_snapshot`, `controller_expand_volume`
+- `success`: `true`, `false`
+- `protocol`: `SMB`, `NFS`
+- `storage_account_type`: `Premium_LRS`, `Premium_ZRS`, `Standard_LRS`, `StandardV2_LRS`, `Standard_GRS`, `Standard_ZRS`, etc.
+
+### Node Metrics (port 29615)
+
+| Metric | Type | Labels | Description |
+|--------|------|--------|-------------|
+| `azurefile_csi_driver_operation_duration_seconds` | Histogram | `operation`, `success` | Duration of CSI operations in seconds |
+| `azurefile_csi_driver_operations_total` | Counter | `operation`, `success` | Total number of CSI operations |
+
+**Label Values:**
+- `operation`: `node_stage_volume`, `node_unstage_volume`, `node_publish_volume`, `node_unpublish_volume`
+- `success`: `true`, `false`
+
+### Azure Cloud Provider Metrics
+
+The CSI driver also exposes Azure cloud provider metrics from the underlying Azure SDK operations:
+
+| Metric | Type | Labels | Description |
+|--------|------|--------|-------------|
+| `cloudprovider_azure_api_request_duration_seconds` | Histogram | `request`, `resource_group`, `subscription_id`, `source`, `result` | Latency of Azure API calls |
+| `cloudprovider_azure_api_request_throttled_count` | Counter | `request`, `resource_group`, `subscription_id`, `source` | Number of throttled Azure API requests |
+| `cloudprovider_azure_api_request_errors` | Counter | `request`, `resource_group`, `subscription_id`, `source` | Number of errors in Azure API requests |
+
+These metrics help monitor Azure API performance, throttling, and error rates for file share operations.
+
 ## Get Prometheus metrics from CSI driver node pod
 
 ```console
-kubectl get --raw /api/v1/namespaces/kube-system/pods/csi-azurefile-node-hfgrn:29615/proxy/metrics
+kubectl get --raw /api/v1/namespaces/kube-system/pods/csi-azurefile-node-xxxxx:29615/proxy/metrics
 ```
+
+> **Note:** Replace `csi-azurefile-node-xxxxx` with an actual pod name from `kubectl get pods -n kube-system -l app=csi-azurefile-node`