Use the Amazon Management Pack
You can use the Amazon Management Pack to monitor the health of your Amazon resources.
Views
The Amazon Management Pack provides the following views, which are displayed in the Monitoring workspace of the Operations console.
Views
EC2 Instances
View the health state of the EC2 instances for a particular Amazon Web Services account, from all Availability Zones and Regions. The view also includes EC2 instances running in a virtual private cloud (VPC). The Amazon Management Pack retrieves tags, so you can search and filter the list using those tags.
When you select an Amazon EC2 instance, you can perform instance health tasks:
-
Open Amazon Console: Launches the Amazon Web Services Management Console in a web browser.
-
Open RDP to Amazon EC2 Instance: Opens an RDP connection to the selected Windows instance.
-
Reboot Amazon EC2 Instance: Reboots the selected EC2 instance.
-
Start Amazon EC2 Instance: Starts the selected EC2 instance.
-
Stop Amazon EC2 Instance: Stops the selected EC2 instance.
EC2 Instances Diagram View
Shows the relationship of an instance with other components.
Amazon EBS Volumes
Shows the health state of all the Amazon EBS volumes for a particular Amazon Web Services account from all Availability Zones and Regions.
Amazon EBS Volumes Diagram View
Shows an Amazon EBS volume and any associated alarms. The following illustration shows an example:
Classic Load Balancers
Shows the health state of all of the Classic Load Balancers for a particular Amazon Web Services account from all Regions.
Elastic Load Balancing Diagram View
Shows the Elastic Load Balancing relationship with other components. The following illustration shows an example:
Amazon Elastic Beanstalk applications
Shows the state of all discovered Amazon Elastic Beanstalk applications.
Amazon Elastic Beanstalk Applications Diagram View
Shows the Amazon Elastic Beanstalk application, application environment, application configuration, and application resources objects.
Amazon CloudFormation stacks
Shows the health state of all the Amazon CloudFormation stacks for a particular Amazon account from all Regions.
Amazon CloudFormation stacks diagram view
Shows the Amazon CloudFormation stack relationship with other components. An Amazon CloudFormation stack might contain Amazon EC2 or Elastic Load Balancing resources. The following illustration shows an example:
Amazon performance views
Shows the Amazon CloudWatch metrics for Amazon EC2, Amazon EBS, and Elastic Load Balancing, custom metrics, and metrics created from CloudWatch alarms. In addition, there are separate performance views for each resource. The Other Metrics performance view contains custom metrics, and metrics created from CloudWatch alarms. For more information about these metrics, see Amazon Services That Publish CloudWatch Metrics in the Amazon CloudWatch User Guide. The following illustration shows an example.
Amazon CloudWatch metric alarms
Shows Amazon CloudWatch alarms related to the discovered Amazon resources.
Amazon alerts
Shows the alerts that the Amazon management pack produces when the health of an object is in a critical state.
Watcher nodes (System Center Operations Manager 2007 R2)
View the health state of the watcher nodes across all of the Amazon Web Services accounts that are being monitored. A Healthy state means that the watcher node is configured correctly and can communicate with Amazon.
Discoveries
Discoveries are the Amazon resources that are monitored by the Amazon Management Pack. The Amazon Management Pack discovers the following objects:
-
Amazon EC2 instances
-
EBS volumes
-
ELB load balancers
-
Amazon CloudFormation stacks
-
Amazon CloudWatch alarms
-
Amazon Elastic Beanstalk applications
-
Amazon EC2 Auto Scaling groups and Availability Zones
Amazon CloudWatch metrics are generated for the following resources:
-
Amazon EC2 instance
-
EBS volume
-
Elastic Load Balancing
-
Custom Amazon CloudWatch metrics
-
Metrics from existing Amazon CloudWatch alarms
For Amazon CloudWatch metrics discovery, the following guidelines apply:
-
Amazon CloudFormation stacks do not have any default Amazon CloudWatch metrics.
-
Stopped Amazon EC2 instances or unused Amazon EBS volumes do not generate data for their default Amazon CloudWatch metrics.
-
After starting an Amazon EC2 instance, it can take up to 30 minutes for the Amazon CloudWatch metrics to appear in Operations Manager.
-
Amazon CloudWatch retains the monitoring data for two weeks, even if your Amazon resources have been terminated. This data appears in Operations Manager.
-
An existing Amazon CloudWatch alarm for a resource that is not supported will create a metric and be associated with the Amazon CloudWatch alarm. These metric can be viewed in the Other Metrics performance view.
The Amazon Management Pack also discovers the following relationships:
-
Amazon CloudFormation stack and its Elastic Load Balancing or Amazon EC2 resources
-
Elastic Load Balancing load balancer and its EC2 instances
-
Amazon EC2 instance and its EBS volumes
-
Amazon EC2 instance and its operating system
-
Amazon Elastic Beanstalk application and its environment, configuration, and resources
The Amazon Management Pack automatically discovers the relationship between an EC2 instance and the operating system running on it. To discover this relationship, the Operations Manager Agent must be installed and configured on the instance and the corresponding operating system management pack must be imported in Operations Manager.
Discoveries run on the management servers in the resource pool (System Center 2012) or the watcher node (System Center 2007 R2).
| Discovery | Interval (seconds) |
|---|---|
|
Amazon Resources Discovery (SCOM 2012) Discovers EC2 instances, Amazon EBS volumes, load balancers, and CloudFront stacks. |
14400 |
|
Amazon Elastic Beanstalk Discovery Discovers Amazon Elastic Beanstalk and its relationship with environment, resources, and configuration. |
14400 |
|
CloudWatch Alarms Discovery Discovers alarms generated using CloudWatch metrics. |
900 |
|
Custom CloudWatch Metric Discovery Discovers custom CloudWatch metrics. |
14400 |
|
Watcher Node Discovery (SCOM 2007 R2) Targets the root management server and creates the watcher node objects. |
14400 |
Monitors
Monitors are used to measure the health of your Amazon resources. Monitors run on the management servers in the resource pool (System Center 2012) or the watcher node (System Center 2007 R2).
| Monitor | Interval (seconds) |
|---|---|
|
Amazon CloudFormation Stack Status |
900 |
|
Amazon CloudWatch Metric Alarm |
300 |
|
Amazon EBS Volume Status |
900 |
|
Amazon EC2 Instance Status |
900 |
|
Amazon EC2 Instance System Status |
900 |
|
Amazon Elastic Beanstalk Status |
900 |
|
Watcher Node to Amazon Cloud Connectivity (SCOM 2007 R2) |
900 |
Rules
Rules create alerts (based on Amazon CloudWatch metrics) and collect data for analysis and reporting.
| Rule | Interval (seconds) |
|---|---|
|
Amazon Resource Discovery Rule (SCOM 2007 R2) Targets the watcher node and uses the Amazon API to discover objects for the following Amazon resources: EC2 instances, EBS volumes, load balancers, and Amazon CloudFormation stacks. (CloudWatch metrics or alarms are not discovered). After discovery is complete, view the objects in the Not Monitored state. |
14400 |
|
Amazon Elastic Block Store Volume Performance Metrics Data Collection Rule |
900 |
|
Amazon EC2 Instance Performance Metrics Data Collection Rule |
900 |
|
Elastic Load Balancing Balancing Performance Metrics Data Collection Rule |
900 |
|
Custom CloudWatch Metric Data Collection Rule |
900 |
Events
Events report on activities that involve the monitored resources. Events are written to the Operations Manager event log.
| Event ID | Description |
|---|---|
|
4101 |
Amazon EC2 Instance Discovery (General Discovery) finished |
|
4102 |
Elastic Load Balancing Metrics Discovery, Amazon EBS Volume Metrics Discovery, Amazon EC2 Instance Metrics Discovery finished |
|
4103 |
Amazon CloudWatch Metric Alarms Discovery finished |
|
4104 |
Amazon Windows Computer Discovery finished |
|
4105 |
Collecting Amazon Metrics Alarm finished |
|
4106 |
EC2 Instance Computer Relation Discovery finished |
|
4107 |
Collecting Amazon CloudFormation Stack State finished |
|
4108 |
Collecting Watcher Node Availability State finished |
|
4109 |
Amazon Metrics Collection Rule finished |
|
4110 |
Task to change Amazon Instance State finished |
|
4111 |
EC2 Instance Status Monitor State finished |
|
4112 |
Amazon EBS Volume Status Monitor State finished |
|
4113 |
Amazon EC2 Instance Scheduled Events Monitor State calculated |
|
4114 |
Amazon EBS Scheduled Events Monitor State calculated |
|
4115 |
Elastic Beanstalk Discovery finished |
|
4116 |
Elastic Beanstalk Environment Status State calculated |
|
4117 |
Elastic Beanstalk Environment Operational State calculated |
|
4118 |
Elastic Beanstalk Environment Configuration State calculated |
Health model
The following illustration shows the health model defined by the Amazon Management Pack.
The health state for a CloudWatch alarm is rolled up to its corresponding CloudWatch metric. The health state for a CloudWatch metric for Amazon EC2 is rolled up to the EC2 instance. Similarly, the health state for the CloudWatch metrics for Amazon EBS is rolled up to the Amazon EBS volume. The health states for the Amazon EBS volumes used by an EC2 instance are rolled up to the EC2 instance.
When the relationship between an EC2 instance and its operating system has been discovered, the operating system health state is rolled up to the EC2 instance.
The health state of an Amazon CloudFormation stack depends on the status of the Amazon CloudFormation stack itself and the health states of its resources, namely the load balancers and EC2 instances.
The following table illustrates how the status of the Amazon CloudFormation stack corresponds to its health state.
| Health State | Amazon CloudFormation Stack Status | Notes |
|---|---|---|
|
Error |
|
Most likely usable |
|
Warning |
|
Recovering after some problem |
|
Healthy |
|
Usable |
The full health model for an Amazon CloudFormation stack is as follows:
Customize the Amazon Management Pack
To change the frequency of discoveries, rules, and monitors, you can override the interval time (in seconds).
To change frequency
-
In the Operations Manager toolbar, click Go, and then click Authoring.
-
In the Authoring pane, expand Management Pack Objects and then click the object to change (for example, Object Discoveries, Rules, or Monitors).
-
In the toolbar, click Scope.
-
In the Scope Management Pack Objects dialog box, click View all targets.
-
To limit the scope to Amazon objects, type Amazon in the Look for field.
-
Select the object want to configure and click OK.
-
In the Operations Manager center pane, right-click the object to configure, click Overrides, and then click the type of override you want to configure.
-
Use the Override Properties dialog box to configure different values and settings for objects.
Tip
To disable a discovery, rule, or monitoring object right-click the object to disable in the Operations Manager center pane, click Overrides, and then click Disable the Rule. You might disable rules if, for example, you do not run Amazon Elastic Beanstalk applications or use custom Amazon CloudWatch metrics.
For information about creating overrides, see Tuning Monitoring by
Using Targeting and Overrides
For information about creating custom rules and monitors, see Authoring for System
Center 2012 - Operations Manager