feat: add windows recording rules to azure monitor metrics addon (managed prometheus)

linter_exclusions.yml

@@ -174,6 +174,9 @@ aks update:
     ksm_metric_labels_allow_list:
       rule_exclusions:
       - option_length_too_long
+    enable_windows_recording_rules:
+      rule_exclusions:
+      - option_length_too_long
     custom_ca_trust_certificates:
       rule_exclusions:
         - option_length_too_long

src/aks-preview/HISTORY.rst

@@ -12,7 +12,10 @@ To release a new version, please select a new version number (usually plus 1 to
 Pending
 +++++++
 
+0.5.132
++++++++
 * Change the short name of option `--source-resource-id` in command `az aks trustedaccess rolebinding create` from `-s` to `-r`.
+* Add parameter to enable windows recording rules `--enable-windows-recording-rules` for the Azure Monitor Metrics addon
 
 0.5.131
 +++++++

src/aks-preview/azext_aks_preview/_help.py

@@ -907,6 +907,9 @@
         - name: --grafana-resource-id
           type: string
           short-summary: Resource ID of the Azure Managed Grafana Workspace
+        - name: --enable-windows-recording-rules
+          type: bool
+          short-summary: Enable Windows Recording Rules when enabling the Azure Monitor Metrics addon
         - name: --disable-azuremonitormetrics
           type: bool
           short-summary: Disable Azure Monitor Metrics Profile

src/aks-preview/azext_aks_preview/_params.py

@@ -502,6 +502,7 @@ def load_arguments(self, _):
         c.argument('ksm_metric_labels_allow_list', validator=validate_ksm_labels, is_preview=True)
         c.argument('ksm_metric_annotations_allow_list', validator=validate_ksm_annotations, is_preview=True)
         c.argument('grafana_resource_id', validator=validate_grafanaresourceid, is_preview=True)
+        c.argument('enable_windows_recording_rules', action='store_true', is_preview=True)
         c.argument('disable_azuremonitormetrics', action='store_true', is_preview=True)
         c.argument('enable_vpa', action='store_true', is_preview=True, help="enable vertical pod autoscaler for cluster")
         c.argument('disable_vpa', action='store_true', is_preview=True, help="disable vertical pod autoscaler for cluster")

src/aks-preview/azext_aks_preview/azuremonitorprofile.py

@@ -19,7 +19,6 @@
 from six import with_metaclass
 from azure.core import CaseInsensitiveEnumMeta
 from azure.core.exceptions import HttpResponseError
-from azure.cli.core.util import get_az_user_agent
 
 AKS_CLUSTER_API = "2022-07-02-preview"
 MAC_API = "2021-06-03-preview"
@@ -508,20 +507,8 @@ def link_grafana_instance(cmd, raw_parameters, azure_monitor_workspace_resource_
     return GrafanaLink.SUCCESS
 
 
-def create_rules(cmd, cluster_subscription, cluster_resource_group_name, cluster_name, azure_monitor_workspace_resource_id, mac_region):
+def put_rules(cmd, default_rule_group_id, default_rule_group_name, mac_region, azure_monitor_workspace_resource_id, cluster_name, default_rules_template, url, i):
     from azure.cli.core.util import send_raw_request
-    with urllib.request.urlopen("https://defaultrulessc.blob.core.windows.net/defaultrules/ManagedPrometheusDefaultRecordingRules.json") as url:
-        default_rules_template = json.loads(url.read().decode())
-    default_rule_group_name = "NodeRecordingRulesRuleGroup-{0}".format(cluster_name)
-    default_rule_group_id = "/subscriptions/{0}/resourceGroups/{1}/providers/Microsoft.AlertsManagement/prometheusRuleGroups/{2}".format(
-        cluster_subscription,
-        cluster_resource_group_name,
-        default_rule_group_name
-    )
-    url = "https://management.azure.com{0}?api-version={1}".format(
-        default_rule_group_id,
-        RULES_API
-    )
     body = json.dumps({
         "id": default_rule_group_id,
         "name": default_rule_group_name,
@@ -531,14 +518,15 @@ def create_rules(cmd, cluster_subscription, cluster_resource_group_name, cluster
             "scopes": [
                 azure_monitor_workspace_resource_id
             ],
+            "enabled": True,
             "clusterName": cluster_name,
             "interval": "PT1M",
-            "rules": default_rules_template["resources"][0]["properties"]["rules"]
+            "rules": default_rules_template["resources"][i]["properties"]["rules"]
         }
     })
     for _ in range(3):
         try:
-            headers = ['User-Agent=azuremonitormetrics.create_rules_node']
+            headers = ['User-Agent=azuremonitormetrics.put_rules.' + default_rule_group_name]
             send_raw_request(cmd.cli_ctx, "PUT", url,
                              body=body, headers=headers)
             error = None
@@ -547,6 +535,24 @@ def create_rules(cmd, cluster_subscription, cluster_resource_group_name, cluster
             error = e
     else:
         raise error
+
+
+def create_rules(cmd, cluster_subscription, cluster_resource_group_name, cluster_name, azure_monitor_workspace_resource_id, mac_region, raw_parameters):
+    from azure.cli.core.util import send_raw_request
+    with urllib.request.urlopen("https://defaultrulessc.blob.core.windows.net/defaultrules/ManagedPrometheusDefaultRecordingRules.json") as url:
+        default_rules_template = json.loads(url.read().decode())
+    default_rule_group_name = "NodeRecordingRulesRuleGroup-{0}".format(cluster_name)
+    default_rule_group_id = "/subscriptions/{0}/resourceGroups/{1}/providers/Microsoft.AlertsManagement/prometheusRuleGroups/{2}".format(
+        cluster_subscription,
+        cluster_resource_group_name,
+        default_rule_group_name
+    )
+    url = "https://management.azure.com{0}?api-version={1}".format(
+        default_rule_group_id,
+        RULES_API
+    )
+    put_rules(cmd, default_rule_group_id, default_rule_group_name, mac_region, azure_monitor_workspace_resource_id, cluster_name, default_rules_template, url, 0)
+
     default_rule_group_name = "KubernetesRecordingRulesRuleGroup-{0}".format(cluster_name)
     default_rule_group_id = "/subscriptions/{0}/resourceGroups/{1}/providers/Microsoft.AlertsManagement/prometheusRuleGroups/{2}".format(
         cluster_subscription,
@@ -557,31 +563,34 @@ def create_rules(cmd, cluster_subscription, cluster_resource_group_name, cluster
         default_rule_group_id,
         RULES_API
     )
-    body = json.dumps({
-        "id": default_rule_group_id,
-        "name": default_rule_group_name,
-        "type": "Microsoft.AlertsManagement/prometheusRuleGroups",
-        "location": mac_region,
-        "properties": {
-            "scopes": [
-                azure_monitor_workspace_resource_id
-            ],
-            "clusterName": cluster_name,
-            "rules": default_rules_template["resources"][1]["properties"]["rules"]
-        }
-    })
-    for _ in range(3):
-        try:
-            headers = ['User-Agent=azuremonitormetrics.create_rules_kubernetes']
-            send_raw_request(cmd.cli_ctx, "PUT", url,
-                             body=body, headers=headers)
-            error = None
-            break
-        except CLIError as e:
-            print(e)
-            error = e
-    else:
-        raise error
+    put_rules(cmd, default_rule_group_id, default_rule_group_name, mac_region, azure_monitor_workspace_resource_id, cluster_name, default_rules_template, url, 1)
+
+    enable_windows_recording_rules = raw_parameters.get("enable_windows_recording_rules")
+
+    if enable_windows_recording_rules is True:
+        default_rule_group_name = "NodeRecordingRulesRuleGroup-Win-{0}".format(cluster_name)
+        default_rule_group_id = "/subscriptions/{0}/resourceGroups/{1}/providers/Microsoft.AlertsManagement/prometheusRuleGroups/{2}".format(
+            cluster_subscription,
+            cluster_resource_group_name,
+            default_rule_group_name
+        )
+        url = "https://management.azure.com{0}?api-version={1}".format(
+            default_rule_group_id,
+            RULES_API
+        )
+        put_rules(cmd, default_rule_group_id, default_rule_group_name, mac_region, azure_monitor_workspace_resource_id, cluster_name, default_rules_template, url, 2)
+
+        default_rule_group_name = "NodeAndKubernetesRecordingRulesRuleGroup-Win-{0}".format(cluster_name)
+        default_rule_group_id = "/subscriptions/{0}/resourceGroups/{1}/providers/Microsoft.AlertsManagement/prometheusRuleGroups/{2}".format(
+            cluster_subscription,
+            cluster_resource_group_name,
+            default_rule_group_name
+        )
+        url = "https://management.azure.com{0}?api-version={1}".format(
+            default_rule_group_id,
+            RULES_API
+        )
+        put_rules(cmd, default_rule_group_id, default_rule_group_name, mac_region, azure_monitor_workspace_resource_id, cluster_name, default_rules_template, url, 3)
 
 
 def delete_dcra(cmd, cluster_region, cluster_subscription, cluster_resource_group_name, cluster_name):
@@ -667,7 +676,7 @@ def link_azure_monitor_profile_artifacts(cmd, cluster_subscription, cluster_reso
     # Link grafana
     link_grafana_instance(cmd, raw_parameters, azure_monitor_workspace_resource_id)
     # create recording rules and alerts
-    create_rules(cmd, cluster_subscription, cluster_resource_group_name, cluster_name, azure_monitor_workspace_resource_id, mac_region)
+    create_rules(cmd, cluster_subscription, cluster_resource_group_name, cluster_name, azure_monitor_workspace_resource_id, mac_region, raw_parameters)
 
 
 def unlink_azure_monitor_profile_artifacts(cmd, cluster_subscription, cluster_resource_group_name, cluster_name, cluster_region):

src/aks-preview/azext_aks_preview/custom.py

@@ -845,6 +845,7 @@ def aks_update(
     ksm_metric_labels_allow_list=None,
     ksm_metric_annotations_allow_list=None,
     grafana_resource_id=None,
+    enable_windows_recording_rules=False,
     disable_azuremonitormetrics=False,
     enable_vpa=False,
     disable_vpa=False,