Migrating to Global Event Handlers

Migrating to Global Event Handlers

In this guide, you will:

• You will migrate to Global Event Handlers (GEH) from Regional Event Handler (REH).
• Decommission Regional Event Handler (REH).
• Monitor and troubleshoot the installation process.

Global Event Handlers (GEH) simplify event management by consolidating event handling into a single primary AWS region while enabling coverage across all regions. By migrating to Global Event Handlers, organizations can achieve cost optimization, operational efficiency, and a streamlined approach to event management in AWS.

Prerequisites

Before proceeding, ensure the following:

• Access to an AWS account with permissions to create IAM roles and policies.
• A primary region designated for GEH (default is us-east-1).
• Active CloudTrail trails for capturing events.

Regional vs Global Event Handlers

Regional Event Handlers (REH)

Each AWS region independently handles events using its own infrastructure. Events are processed regionally and sent directly to the Guardrails event ingestion endpoint.

Resources Deployed Per Region

• EventBridge Rules: Filters specific API events.
• EventBridge Targets: Routes events to SNS topics.
• SNS Topics and Subscriptions: Publish and forward events to Guardrails.

Global Event Handlers (GEH)

A designated primary AWS region manages all event processing, while secondary regions forward their events to the primary region for centralized handling. By default, the primary region is set to us-east-1, but this can be customized by configuring the policy AWS > Turbot > Event Handlers [Global] > Primary Region with the desired region value.

Resources Deployed

• IAM Role for Cross-Region Event Forwarding: Enables seamless event forwarding from all GEH secondary regions to the primary region.

• Primary Region

• EventBridge Rules: Define custom event patterns and sources.

• EventBridge Targets: Direct events to an SNS Topic.

• SNS Topics and Subscriptions: Publish and forward events to Guardrails.

• Secondary Regions

• EventBridge Rule: Captures event sources and forwards them to the primary region.

• EventBridge Target: Sends events to the EventBridge bus in the primary region.

Step 1: Deploying the EventBridge IAM Role

To enable seamless data transfer between regional event buses, the EventBridge IAM role is a critical component of the Global Event Handlers (GEH) setup. This role allows secondary regions to forward events to the primary region for centralized processing. GEH will only use the default event bus. There is no need to create second event bus exclusively for GEH.

Options for Creating the EventBridge IAM Role

1. Role Creation by Turbot: Turbot can create the required IAM roles for you. Simply enable the AWS > Turbot > Service Roles control, and Turbot will handle the setup seamlessly.

[!NOTE] The Turbot Service Roles control is designed to configure various roles required for multiple dependent controls. If you are using Turbot Service Roles exclusively for Global Event Handlers, you can disable the other Service Roles that are not needed.

Below is a list of policies to enable or disable:

2. Self-Managed Role Creation: If you prefer to create the role yourself, follow the steps below to define the required Assume Role Policy and IAM Policy.

Assume Role Policy

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Principal": {
        "Service": "events.amazonaws.com"
      },
      "Action": "sts:AssumeRole"
    }
  ]
}

IAM Policy

Attach the following IAM policy to grant necessary permissions to the EventBridge role. This can be a separate policy or an inline policy.

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": ["events:PutEvents"],
      "Resource": "arn:{PARTITION}:events:{GLOBAL_EVENTS_PRIMARY_REGION}:{AWS_ACCOUNT_ID}:event-bus/default"
    }
  ]
}

Terraform Example for the same.

resource "aws_iam_role" "event_handlers_global_role" {
  name = "${var.event_handlers_global_role_name}"
  path = "${var.service_roles_path}"
  assume_role_policy = jsonencode({
    Version = "2012-10-17",
    Statement = [
      {
        Effect    = "Allow",
        Principal = { Service = "events.amazonaws.com" },
        Action    = "sts:AssumeRole"
      }
    ]
  })
}

resource "aws_iam_policy" "turbot_guardrails_geh_eventbridge_policy" {
  name = "${var.event_handlers_global_policy_name}"

  policy = jsonencode({
    Version = "2012-10-17",
    Statement = [
      {
        Effect   = "Allow",
        Action   = ["events:PutEvents"],
        Resource = "arn:${var.partition}:events:${var.primary_region}:${var.account_id}:event-bus/default"
      }
    ]
  })
}

resource "aws_iam_role_policy_attachment" "turbot_guardrails_role_policy_attachment" {
  role       = aws_iam_role.event_handlers_global_role.name
  policy_arn = aws_iam_policy.turbot_guardrails_geh_eventbridge_policy.arn
}

Step 2: Configure the IAM Role ARN Policy Setting for Event Forwarding

If you are using Turbot Service Roles, this step is automatically handled, and no further action is required. However, if you are manually creating the roles, you need to configure the AWS > Turbot > Event Handlers [Global] > Events > Target > IAM Role ARN policy. This can be done through the Turbot UI or using the Terraform configuration provided below.

Terraform Configuration

The following Terraform resource sets the IAM Role ARN policy for EventBridge in Global Event Handlers:

resource "turbot_policy_setting" "aws_event_handlers_global_events_target_iam_role_arn" {
  resource       = "RESOURCE_ID" # or the policy pack that holds the other GEH policies
  type           = "tmod:@turbot/aws#/policy/types/eventHandlersGlobalEventsTargetIamRoleArn"
  template_input = <<EOT
{
  account
  {
    # Look for the name of the EventBridge IAM role.
    event_bridge_role: children(filter:"$.RoleName:${var.EVENTBRIDGE_IAM_ROLE} resourceTypeId:tmod:@turbot/aws-iam#/resource/types/role level:self"){
      role:items
      {
        akas
      }
    }
  }
}
EOT
  template       = <<EOT
{% if $.account.event_bridge_role.role | length == 1 -%}
{{ $.account.event_bridge_role.role[0].akas[0] }}
{% else -%}
{#  if something weird happens, the GEH control will go into `error` because the ARN isn't well-formed. #}
"missing or too many eventbridge roles"
{% endif -%}
EOT
}

Step 3: Deploy Global Event Handlers (GEH)

1. In the Guardrails console navigate to the Policies and search for AWS > Turbot > Event Handlers [Global] policy. Select New Policy Setting
2. Choose Resource as Turbot and Setting as Enforce: Configured
3. Validate the deployment using the Global Event Handler Report link https://{workspace}/apollo/reports/controls-by-state?filter=controlTypeId%3A%27tmod%3A%40turbot%2Faws%23%2Fcontrol%2Ftypes%2FeventHandlersGlobal%27
4. Ensure all GEH controls are in the ok state with the message "All required resources exist".

Step 4: Decommission Regional Event Handlers (EH)

1. Set the AWS > Turbot > Event Handlers policy to Enforce: Not configured to trigger cleanup of legacy event handler infrastructure.

[!IMPORTANT] In the Resource dropdown selection allows you to choose Turbot or specific accounts. If Turbot is selected, then it applies the settings across all the AWS accounts.

[!WARNING] Setting the event handlers to Skip at this point will leave event handler infrastructure deployed concurrent to global event handlers; effectively doubling the events sent back to Guardrails for processing.

2. Validate the cleanup using the Event Handler Report link https://{workspace}/apollo/reports/controls-by-state?filter=controlTypeId%3A%27tmod%3A%40turbot%2Faws%23%2Fcontrol%2Ftypes%2FeventHandlers%27
3. Ensure all AWS > Turbot > Event Handlers controls are in the ok state with the message ""Empty configuration - no action needed".
4. Delete the Event Handler policy settings to complete decommissioning. This will change the Event Handler controls to Skipped.

Step 5: Verify Events

The global event handlers are now configured in the target account. To verify they are functioning correctly:

1. Primary Region Testing: Create a resource in the primary region and verify its detection. Confirm that the associated controls are triggered and executed based on the policies set in the Guardrails console.

2. Secondary Region Testing: Create a resource in a secondary region and verify its detection. Ensure that the associated controls are triggered and executed according to the policies set in the Guardrails console.

3. Event Handler Cleanup: Confirm that legacy EH resources are removed from the Guardrails CMDB.

Troubleshooting