Update support for custom identifier through deployment.

Provide error checking skip platform specific alarms when
the platform can't be identified from image.
Remove default alarm for CPUCreditBalance.

CloudWatchAutoAlarms.yaml

@@ -22,6 +22,11 @@ Parameters:
     Description: Enter the Amazon SNS Notification ARN for alarm notifications, leave blank to disable notifications.
     Type: String
     Default: ""
+  AlarmIdentifierPrefix:
+    Description: Enter the prefix that should be added to the beginning of each alarm created by the solution, (e.g. AutoAlarm-i-00e4f327736cb077f-CPUUtilization-GreaterThanThreshold-80-5m)
+    Type: String
+    Default: AutoAlarm
+
 
 Conditions:
   ConfigureAlarmNotifications: !Not [!Equals ["", !Ref AlarmNotificationARN]]
@@ -48,6 +53,7 @@ Resources:
           ALARM_CPU_CREDIT_BALANCE_LOW_THRESHOLD: 100
           ALARM_MEMORY_HIGH_THRESHOLD: 75
           ALARM_DISK_PERCENT_LOW_THRESHOLD: 20
+          ALARM_IDENTIFIER_PREFIX: !Ref AlarmIdentifierPrefix
           CLOUDWATCH_APPEND_DIMENSIONS: 'InstanceId, ImageId, InstanceType'
 
           ALARM_LAMBDA_ERROR_THRESHOLD: 0
@@ -102,7 +108,7 @@ Resources:
                   - cloudwatch:DescribeAlarms
                   - cloudwatch:DeleteAlarms
                   - cloudwatch:PutMetricAlarm
-                Resource:  !Sub "arn:${AWS::Partition}:cloudwatch:${AWS::Region}:${AWS::AccountId}:alarm:AutoAlarm-*"
+                Resource:  !Sub "arn:${AWS::Partition}:cloudwatch:${AWS::Region}:${AWS::AccountId}:alarm:${AlarmIdentifierPrefix}-*"
               - Effect: Allow
                 Action:
                   - cloudwatch:DescribeAlarms

README.md

@@ -66,7 +66,11 @@ The following list provides a description of the setting along with the environm
     * You can add EC2 metric dimensions to all metrics collected by the CloudWatch agent.  This environment variable aligns to your CloudWatch configuration setting for [**append_dimensions**](https://docs.aws.amazon.com/AmazonCloudWatch/latest/monitoring/CloudWatch-Agent-Configuration-File-Details.html#CloudWatch-Agent-Configuration-File-Metricssection).  The default setting includes all the supported dimensions:  InstanceId, ImageId, InstanceType, AutoScalingGroupName
 * **DEFAULT_ALARM_SNS_TOPIC_ARN**:  arn:${AWS::Partition}:sns:${AWS::Region}:${AWS::AccountId}:CloudWatchAutoAlarmsSNSTopic
     * You can define an Amazon Simple Notification Service (Amazon SNS) topic that the Lambda function will specify as the notification target for created alarms. You provide the Amazon SNS Topic Amazon Resource Name (ARN) with the **AlarmNotificationARN** parameter when you deploy the CloudWatchAutoAlarms.yaml CloudFormation template.  If you leave the **AlarmNotificationARN** parameter value blank, then this environment variable is not set and created alarms won't use notifications.
-* You can update the thresholds for the default alarms by updating the following environment variables:
+* **ALARM_IDENTIFIER_PREFIX**:  AutoAlarm
+    * The prefix name that is added to the beginning of each CloudWatch alarm created by the solution.  (e.g. For "AutoAlarm":  (e.g. AutoAlarm-i-00e4f327736cb077f-CPUUtilization-GreaterThanThreshold-80-5m))  You should update this variable via the **AlarmIdentifierPrefix** in the [CloudWatchAutoAlarms.yaml](./CloudWatchAutoAlarms.yaml) CloudFormation template so that the IAM policy is updated to align with your custom name. 
+
+You can update the thresholds for the default alarms by updating the following environment variables:
+
 
    **For Amazon EC2**:
     * **ALARM_CPU_HIGH_THRESHOLD**: 75
@@ -135,6 +139,15 @@ In order to create the default alarm set for an Amazon EC2 instance or AWS Lambd
 
 For Amazon EC2 instances, you must add this tag during instance launch or you can add this tag at any time to an instance and then stop and start the instance in order to create the default alarm set as well as any custom, instance specific alarms.
 
+You can also manually invoke the CloudWatchAutoAlarms lambda function with the following event payload to create / update EC2 alarms without having to stop and start your EC2 instances:
+
+```json
+{
+  "action": "scan"
+}
+```
+You can do this with a test execution of the CloudWatchAUtoAlarms AWS Lambda function.  Open the  AWS Lambda Management Console and perform a test invocation from the **Test** tab with the payload provided here.
+
 For AWS Lambda, you can add this tag to an AWS Lambda function at any time in order to create the default alarm set as well as any custom, function specific alarms.
 
 

src/actions.py

@@ -220,9 +220,11 @@ def process_alarm_tags(instance_id, instance_info, default_alarms, metric_dimens
     if create_default_alarms_flag == 'true':
         for alarm_tag in default_alarms['AWS/EC2']:
             create_alarm_from_tag(instance_id, alarm_tag, instance_info, metric_dimensions_map, sns_topic_arn, alarm_separator, alarm_identifier)
-
-        for alarm_tag in default_alarms[cw_namespace][platform]:
-            create_alarm_from_tag(instance_id, alarm_tag, instance_info, metric_dimensions_map, sns_topic_arn, alarm_separator, alarm_identifier)
+        if platform:
+            for alarm_tag in default_alarms[cw_namespace][platform]:
+                create_alarm_from_tag(instance_id, alarm_tag, instance_info, metric_dimensions_map, sns_topic_arn, alarm_separator, alarm_identifier)
+        else:
+            logger.warning("Skipping platform specific alarm creation for {}, unknown platform.".format(instance_id))
     else:
         logger.info("Default alarm creation is turned off")
 

src/cw_auto_alarms.py

@@ -3,6 +3,9 @@
 from os import getenv
 
 logger = logging.getLogger()
+log_level = getenv("LOGLEVEL", "INFO")
+level = logging.getLevelName(log_level)
+logger.setLevel(level)
 
 create_alarm_tag = getenv("ALARM_TAG", "Create_Auto_Alarms")
 
@@ -14,7 +17,6 @@
 append_dimensions = [dimension.strip() for dimension in append_dimensions.split(',')]
 
 alarm_cpu_high_default_threshold = getenv("ALARM_CPU_HIGH_THRESHOLD", "75")
-alarm_credit_balance_low_default_threshold = getenv("ALARM_CPU_CREDIT_BALANCE_LOW_THRESHOLD", "100")
 alarm_memory_high_default_threshold = getenv("ALARM_MEMORY_HIGH_THRESHOLD", "75")
 alarm_disk_space_percent_free_threshold = getenv("ALARM_DISK_PERCENT_LOW_THRESHOLD", "20")
 alarm_disk_used_percent_threshold = 100 - int(alarm_disk_space_percent_free_threshold)
@@ -27,7 +29,7 @@
 sns_topic_arn = getenv("DEFAULT_ALARM_SNS_TOPIC_ARN", None)
 
 alarm_separator = '-'
-alarm_identifier = 'AutoAlarm'
+alarm_identifier = getenv("ALARM_IDENTIFIER_PREFIX", 'AutoAlarm')
 # For Redhat, the default device is xvda2, xfs, for Ubuntu, the default fstype is ext4,
 # for Amazon Linux, the default device is xvda1, xfs
 default_alarms = {
@@ -38,11 +40,6 @@
             'Key': alarm_separator.join(
                 [alarm_identifier, 'AWS/EC2', 'CPUUtilization', 'GreaterThanThreshold', '5m', 'Average', 'default1']),
             'Value': alarm_cpu_high_default_threshold
-        },
-        {
-            'Key': alarm_separator.join(
-                [alarm_identifier, 'AWS/EC2', 'CPUCreditBalance', 'LessThanThreshold', '5m', 'Average', 'default1']),
-            'Value': alarm_credit_balance_low_default_threshold
         }
     ],
     'AWS/Lambda': [