Sending alerts to Amazon EventBridge

In Mackerel, JSON of alert contents can be received with Amazon EventBridge. This can be useful for tasks such as using AWS Lambda etc. to automate post-notification processing and more.

Adding Amazon EventBridge notifications

Configuration from the console

Notification settings can be added using the Amazon EventBridge form on the Channels page. Specify your AWS account ID, the AWS region that will receive the notifications, and an event name to identify the notifications.

  • For how to check your AWS account ID, please refer to AWS Account Identifiers - AWS General Reference.
  • Mackerel's event source name with Amazon EventBridge follows the format below.   - aws.partner/{organization name}/{customer specified event name}

Integrated configuration from the AWS console

Once you've completed the configuration in Mackerel, you'll need to create an event bus/rule with Amazon EventBridge in order to receive notifications.

  1. Open the Partner Event Sources page from the AWS console.
  2. Switch the console's region of operation to the AWS region to receive notifications that was set in Mackerel.
  3. Select the added event source and press the 'Associate with event bus' button in the upper right.
  4. Select the necessary access permission and click the 'Associate' button.
  5. Open the Rules page from the leftside panel, select the event bus that you created from the drop-down list, and click 'Create rule' at the bottom.
  6. In the Create rule screen, set the name, pattern, and targets. Then, in the Select event bus section, select the 'Custom or partner event bus' option and set your event bus.

JSON format

The JSON that will be sent contains the following. (Items may be added at any time)

  "orgName": "Macker...",
  "event": "alert",
  "memo": "memo....",
  "host": {
    "id": "22D4...",
    "name": "app01",
    "url": "",
    "type": "unknown",
    "status": "working",
    "memo": "",
    "isRetired": false,
    "roles": [
        "fullname": "Service: Role",
        "serviceName": "Service",
        "serviceUrl": "",
        "roleName": "Role",
        "roleUrl": ""
  "alert": {
    "openedAt": 1473129912,
    "closedAt": 1473130092,
    "createdAt": 1473129912693, 
    "criticalThreshold": 1.9588528112516932, 
    "duration": 5, 
    "isOpen": true, 
    "metricLabel": "MetricName", 
    "metricValue": 2.255356387321597, 
    "monitorName": "MonitorName", 
    "monitorOperator": ">", 
    "status": "critical", 
    "trigger": "monitor", 
    "id": "2bj...",
    "url": "",
    "warningThreshold": 1.4665636369580741
  "resourceInfo": {

JSON items

orgName string name of the organization where the alert occurred
event string fixed as alert
memo string notes for the monitoring configuration.
host object host information (output for alerts of host metrics only)
service object service information (output for alerts of service metrics only)
alert object alert information
resourceInfo object AWS resource information (output for supported components only)

JSON items (host information)

id string host ID
name string host name
url string URL of the host details
status string host status (working, standby, poweroff, maintenance)
memo string memo registered to the host
isRetired boolean whether or not retired
roles array[object] role information (if the role is configured to the host)

JSON items (role information)

fullname string service name: role name
serviceName string service name
serviceUrl string URL of the service details
roleName string role name
roleUrl string URL of the role in service details

JSON items (service information)

id string service ID
memo string memo registered to services
name string service name
orgId string ID of the organization where the service is registered
roles array[object] role information (if the role is registered in services)

JSON Items (Metric information by query monitoring)

labels object labels of the metric e.g. { "http.method": "GET", "http.status_code": "200" }
name string metric name e.g. httpcheck.status

JSON items (alert information)

id string alert ID
status string alert status (ok, warning, critical, unknown)
isOpen boolean whether or not the alert is open
trigger string what triggered the notification (monitoring, manual, monitorDelete, hostRetire)
url string URL of the alert details
openedAt number time the alert occurred (epoch seconds)
closedAt number time the alert resolved (epoch seconds)
createdAt number time the alert occurred (epoch milliseconds). A deprecated item and not recommended. Please use openedAt
monitorName string name of the monitoring item that detected the alert
metricLabel string title of the monitoring target’s metrics etc.
metricValue number metric value at the time of detection
criticalThreshold number CRITICAL threshold
warningThreshold number WARNING threshold
monitorOperator string > or <
duration number monitoring interval

※Alert information items fluctuate depending on the type of monitoring target (host metrics, service metrics, external monitoring, expression monitoring, query monitoring) and each configuration item. For example, if only the URL is configured with external monitoring, information for metrics, thresholds, etc. will not be included. The type of monitoring target and the output items correspond with possible input items from the monitoring configurations screen.

JSON items (AWS resource information)


When an alert occurs in an EC2 instance installed with mackerel-agent or AWS integration, the information that has already been acquired by Mackerel*1 will be returned with each field corresponding to ARN format.

service string [optional] service name ec2
region string [optional] region name
accountId string [optional] AWS account ID
resourceType string [optional] resource type instance
resourceId string [optional] resource ID (EC2 instance ID) i-12345...

*1 However, you cannot retrieve this information if cloud_platform is set to none in the mackerel-agent settings.