Feature/dynamic alerts

ingredient/ingredient-metric-alert/README.md

@@ -8,26 +8,19 @@ The Metric Alert ingredient is a plugin for Bake.  When included in a recipe, th
 
 ### Recipe
 ```yaml
-#Provide name 
 name: Event Hub Namespace for diagnostics with a custom metric alert
 shortName: diagnostics
 version: 0.0.1
-#Specify the names of the ingredients to use in the recipe.  This is the name of the ingredient in package.json.  
-#Specify the local path to the module during development.
 ingredients:
   - "@azbake/ingredient-metric-alert@~0"
-  - "@azbake/ingredient-arm@~0"
   - "@azbake/ingredient-event-hub-namespace@~0"
-#Deploys to regions in parallel.  Typically true unless the sequence of deploying to regions is important.
 parallelRegions: true
-#
-#rgOverride: 
+#rgOverride:
 resourceGroup: true
 variables:
 recipe:
   ehndiag-deploy: 
     properties:
-    #Specify the Bake ingredient above
       type: "@azbake/ingredient-event-hub-namespace"
       source: ""
       parameters:    
@@ -37,10 +30,8 @@ recipe:
         skuCapacity: "1"
         isAutoInflateEnabled: "true"
         maximumThroughputUnits: "10"  
-  #Name the deployment.  This shows up in the log window and is the name of the deployment within Azure.
   alert-deploy: 
     properties:
-    #Specify the Bake ingredient above
       type: "@azbake/ingredient-metric-alert"
       source: "[eventhubnamespace.get_resource_profile()]"
       parameters:
@@ -54,33 +45,60 @@ recipe:
         threshold: "0"
         timeAggregation: "Maximum"
         windowSize: "PT5M"
-        evaluationFrequency: "PT1M"
+        evaluationFrequency: "PT5M"
         actionGroupName: "[coreutils.create_resource_name('act','emailops', false)]"
-        actionGroupResourceGroup: "[coreutils.create_resource_name('','actiongroups', false)]"
+        actionGroupResourceGroup: "[coreutils.create_resource_name('','actiongroups', true)]"
         #Uncomment below and comment above if using rgOverride 
         #actionGroupResourceGroup: "[coreutils.resource_group()]"
+        #alertType: "Static"
+        alertType: "Dynamic"
+        #Dynamic alert params have defaults.  Uncomment and specify below to override defaults.
+        # dynamicAlertParameters:
+        #   alertSensitivity: "Low"
+        #   failingPeriods:
+        #     numberOfEvaluationPeriods: "4"
+        #     minFailingPeriodsToAlert: "3"
     dependsOn:
       - ehndiag-deploy
 ```
 
 | property|required|description|
 |---------|--------|-----------|
-| source | yes | Name of the target resource for the metric that the alert is monitoring |
-| alertName | yes | Name of the Alert resource |
-| alertSeverity | no | The severity level of the alert.  Defaults to 3. |
-| isEnabled | no | Specifies whether the alert is enabled.  Defaults to true. |
+| source | yes | Name of the target resource for the metric that the alert is monitoring. |
+| alertName | yes | Name of the Alert resource. |
+| alertSeverity | no (default `3`) | The severity level of the alert. |
+| isEnabled | no (default `true`) | Specifies whether the alert is enabled.|
 | source-type | yes | The [metric namespace](https://docs.microsoft.com/en-us/azure/azure-monitor/platform/metrics-supported) of the alert.  An example is Microsoft.KeyVault/vaults.  |
-| metricName | yes | The name of the metric that the alert is monitoring |
-| operator | no | The comparison operator for the alert.  Defaults to "GreaterThan". |
+| metricName | yes | The name of the metric that the alert is monitoring. |
+| operator | yes | The comparison operator for the alert. |
 | threshold | yes | The threshold value to trigger the alert. |
-| timeAggregation | yes | The aggregation function used for comparing against the threshold.  Defaults to "Average".|
-| windowSize | no | Period of time used to monitor alert activity based on the threshold.  Defaults to PT5M. Must be between five minutes and one day. ISO 8601 duration format. |
-| evaluationFrequency| no | How often the metric alert is evaluated represented in ISO 8601 duration format.  Defaults to PT1M. |
+| timeAggregation | no (default `Average`) | The aggregation function used for comparing against the threshold. 
+| windowSize | no (default `PT5M`) | Period of time used to monitor alert activity based on the threshold.  Must be between five minutes and one day. ISO 8601 duration format. |
+| evaluationFrequency| no (default `PT5M`) | How often the metric alert is evaluated represented in ISO 8601 duration format.|
 | actionGroupName | yes | The name of the action group to trigger when the alert is fired.|
 | actionGroupResourceGroup | yes | The action group's resource group.|
+| alertType | yes | Static or Dynamic (AIOps) |
+| dynamicAlertProperties | no (default below) | Object for specifying dynamic alert properties. |
 
+**dynamicAlertProperties object**
+| property|required|description|
+|---------|--------|-----------|
+| alertSensitivity | no (default `Low`) | Low, Medium, and High sensitivities (in order of least to most noisy alerts.)
+| failingPeriods | no (default below) | Object for specifying dynamic alert failing periods properties. |
+
+**failingPeriods object** - this advanced criteria defaults to fire the alert if 3 of of the last 4 evaluation periods failed
+| property|required|description|
+|---------|--------|-----------|
+| numberOfEvaluationPeriods | no (default `4`) | The total number of periods for evaluating against the failing periods. |
+| minFailingPeriodsToAlert | no (default `3`)| The minimum number of failing periods to fire the alert. |
+
+
+**See Azure Monitor documentation for additional details**
+
+[Platform metric alert rules - Using Resource Manager template](https://docs.microsoft.com/en-us/azure/azure-monitor/platform/alerts-metric-create-templates)
+
+[Azure Monitor REST API - Metric Alerts - Create or Update](https://docs.microsoft.com/en-us/rest/api/monitor/metricalerts/createorupdate)
 
-See [Event Hub SDK documentation for additional details](https://docs.microsoft.com/en-us/dotnet/api/microsoft.azure.management.eventhub.models.eventhub?view=azure-dotnet#properties)
 
 ## Utilities
 Utility classes can be used inside of the bake.yaml file for parameter and source values.

ingredient/ingredient-metric-alert/src/staticAlert.json → ingredient/ingredient-metric-alert/src/alert.json

@@ -55,22 +55,22 @@
     },
     "operator": {
       "type": "string",
-      "defaultValue": "GreaterThan",
       "allowedValues": [
         "Equals",
         "NotEquals",
         "GreaterThan",
         "GreaterThanOrEqual",
         "LessThan",
-        "LessThanOrEqual"
+        "LessThanOrEqual",
+        "GreaterOrLessThan"
       ],
       "metadata": {
-        "description": "Operator comparing the current value with the threshold value."
+        "description": "Operator comparing the current value with the threshold.  GreaterOrLessThan, GreaterThan, and LessThan are applicable to dynamic.  All applicable to static except GreaterOrLessThan."
       }
     },
     "threshold": {
       "type": "string",
-      "defaultValue": "0",
+      "defaultValue": "",
       "metadata": {
         "description": "The threshold value at which the alert is activated."
       }
@@ -97,7 +97,7 @@
     },
     "evaluationFrequency": {
       "type": "string",
-      "defaultValue": "PT1M",
+      "defaultValue": "PT5M",
       "metadata": {
         "description": "How often the metric alert is evaluated represented in ISO 8601 duration format"
       }
@@ -107,6 +107,20 @@
     },
     "actionGroupResourceGroup": {
       "type": "string"      
+    },
+    "alertType": {
+      "type": "string",
+      "allowedValues": ["Static","Dynamic"]
+    },
+    "dynamicAlertProperties": {
+      "type": "object",
+      "defaultValue": {
+        "alertSensitivity": "Low",
+        "failingPeriods": {
+          "numberOfEvaluationPeriods": "4",
+          "minFailingPeriodsToAlert": "3"
+        }
+      }
     }
   },
   "variables": {},
@@ -125,14 +139,17 @@
         "evaluationFrequency": "[parameters('evaluationFrequency')]",
         "windowSize": "[parameters('windowSize')]",
         "criteria": {
-          "odata.type": "Microsoft.Azure.Monitor.SingleResourceMultipleMetricCriteria",
+          "odata.type": "[concat('Microsoft.Azure.Monitor.', if(equals(parameters('alertType'),'Static'),'SingleResourceMultipleMetricCriteria','MultipleResourceMultipleMetricCriteria'))]",
           "allOf": [
             {
+              "criterionType": "[if(equals(parameters('alertType'),'Static'),'StaticThresholdCriterion','DynamicThresholdCriterion')]",
               "name": "1st criterion",
               "metricName": "[parameters('metricName')]",
               "dimensions": [],
               "operator": "[parameters('operator')]",
-              "threshold": "[parameters('threshold')]",
+              "threshold": "[if(equals(parameters('alertType'),'Static'),parameters('threshold'),'')]",
+              "alertSensitivity": "[if(equals(parameters('alertType'),'Dynamic'),parameters('dynamicAlertProperties').alertSensitivity,'')]",
+              "failingPeriods": "[if(equals(parameters('alertType'),'Dynamic'),parameters('dynamicAlertProperties').failingPeriods,json('null'))]",
               "timeAggregation": "[parameters('timeAggregation')]"
             }
           ]

ingredient/ingredient-metric-alert/src/plugin.ts

@@ -1,7 +1,6 @@
 import { BaseIngredient, IngredientManager } from "@azbake/core"
 import { ARMHelper } from "@azbake/arm-helper"
-import staticAlertARMTemplate from "./staticAlert.json"
-import dynamicAlertARMTemplate from "./dynamicAlert.json"
+import alertARMTemplate from "./alert.json"
 
 export class MetricAlertPlugin extends BaseIngredient {
 
@@ -24,18 +23,19 @@ export class MetricAlertPlugin extends BaseIngredient {
             params["source-rg"] = { "value": resource.resourceGroup };
             params["source-name"] = { "value": resource.resource };
 
-            //Generate alertName param as env + region + "alert" + source resource + time aggregation (max, min, etc.) + metric name. 
+            //Generate alertName param as env + region + "alert" + source resource + time aggregation (max, min, etc.) + metric name + alert type (static, dynamic). 
             //Alert names can be up to 128 characaters long in Azure.
-            //Ex) deveusalert-sbwounds-maximum-throttledrequests
+            //Ex) deveusalert-sbwounds-maximum-throttledrequests-dynamic
             const timeAggregation = params["timeAggregation"].value;
             const metricName = params["metricName"].value;
             const sourceName = resource.resource;
-            const tempName = '-' + sourceName + '-' + timeAggregation + '-' + metricName;
-            const alertName = util.create_resource_name("alert", tempName, true);
+            const alertType = params["alertType"].value;
+            const tempName = '-' + sourceName + '-' + timeAggregation + '-' + metricName + '-' + alertType;
+            const alertName = util.create_resource_name("alert", tempName, true);            
             this._logger.log(alertName);
             params["alertName"] = { "value": alertName };
 
-            await helper.DeployTemplate(this._name, staticAlertARMTemplate, params, await util.resource_group())
+            await helper.DeployTemplate(this._name, alertARMTemplate, params, await util.resource_group())
 
         } catch (error) {
             this._logger.error('deployment failed: ' + error)