⚙️ Monitor & auto-heal AWS EC2 instances with multi-channel alerts

⚡ 140 views · ⚙️ DevOps & CI/CD

Description

This n8n workflow automates the monitoring, health assessment, and self-healing of AWS EC2 instances in production environments. It runs periodic checks, identifies unhealthy instances based on status and metrics, restarts them automatically, and notifies teams via multi-channel alerts while logging data for auditing and reporting.

Key Features

• Triggers health checks every 5 minutes to proactively monitor EC2 fleet status.
• Fetches and loops through all production EC2 instances for individualized analysis.
• Evaluates instance health using AWS metrics and custom thresholds to detect issues like high CPU or stopped states.
• Performs automatic restarts on unhealthy instances to minimize downtime.
• Sends instant WhatsApp notifications for urgent alerts, detailed email reports for team review, and logs metrics to Google Sheets for long-term tracking.
• Includes sticky notes for quick reference on configuration, self-healing logic, and alert setup.

Workflow Process

• The Schedule Trigger node runs the workflow every 5 minutes, ensuring frequent health monitoring without overwhelming AWS APIs.
• The Get EC2 Instances node fetches all production-tagged EC2 instances from AWS, filtering by environment (e.g., tag: Environment=Production).
• The Loop Over Instances node iterates through each fetched instance individually, allowing parallel processing for scalability.
• The Check Instance Status node retrieves detailed health metrics for the current instance via AWS API (e.g., status checks, CPU utilization, and state).
• The Health Status Check node evaluates the instance’s status against predefined thresholds (e.g., failed system checks or high load); if healthy, it skips to logging.
• The Analyze Health Data node assesses metrics in depth to determine action (e.g., restart if CPU > 90% for 5+ minutes) and prepares alert payloads.
• The Restart Instance node automatically initiates a reboot on unhealthy instances using AWS EC2 API, with optional dry-run mode for testing.
• The WhatsApp Notification node (part of Multi-Channel Alerts) sends instant alerts via Twilio WhatsApp API, including instance ID, issue summary, and restart status.
• The Email Report node generates and sends a detailed HTML report to the team via SMTP, summarizing checked instances, actions taken, and metrics trends.
• The Google Sheets Logging node appends health data, timestamps, and outcomes to a specified spreadsheet for historical analysis and dashboards.
• The Sticky Notes nodes provide inline documentation: one for AWS credential setup, one explaining self-healing thresholds, and one for alert channel configurations.

Setup Instructions

• Import the workflow into n8n and activate the Schedule Trigger with a 5-minute cron expression (e.g., */5 * * * *).
• Configure AWS credentials in the Get EC2 Instances, Check Instance Status, and Restart Instance nodes using IAM roles with EC2 read/restart permissions.
• Set up Twilio credentials in the WhatsApp Notification node, including your Twilio SID, auth token, and WhatsApp-enabled phone numbers for sender/receiver.
• Add SMTP credentials (e.g., Gmail or AWS SES) in the Email Report node, and update sender/receiver email addresses in the node parameters.
• Link Google Sheets in the Google Sheets Logging node by providing the spreadsheet ID, sheet name, and OAuth credentials for write access.
• Customize health thresholds in Health Status Check and Analyze Health Data (e.g., via expressions for CPU/memory limits).
• Test the workflow by manually executing it on a small set of instances and verifying alerts/logging before enabling production scheduling.
• Review sticky notes within n8n for quick tips, and monitor executions in the dashboard to fine-tune intervals or error handling.

Prerequisites

• AWS account with EC2 access and IAM user/role for DescribeInstances, DescribeInstanceStatus, and RebootInstances actions.
• Twilio account with WhatsApp sandbox or approved number for notifications.
• SMTP email service (e.g., Gmail, Outlook) with app-specific passwords enabled.
• Google Workspace or personal Google account for Sheets integration.
• n8n instance with AWS, Twilio, SMTP, and Google Sheets nodes installed (cloud or self-hosted).
• Production EC2 instances tagged consistently (e.g., Environment=Production) for filtering.

Modification Options

• Adjust the Schedule Trigger interval to hourly for less frequent checks or integrate with AWS CloudWatch Events for dynamic triggering.
• Expand Analyze Health Data to include advanced metrics (e.g., disk I/O via CloudWatch) or ML-based anomaly detection.
• Add more alert channels in Multi-Channel Alerts, such as Slack webhooks or PagerDuty integrations, by duplicating the WhatsApp/Email branches.
• Enhance Google Sheets Logging with charts or conditional formatting via Google Apps Script for visual dashboards.
• Implement approval gates in Restart Instance (e.g., via email confirmation) to prevent auto-restarts in sensitive environments.

Explore More AI Workflows: Get in touch with us for custom n8n automation!

🔗 Nodes Used

Send Email, Google Sheets, HTTP Request, Schedule Trigger

📥 Import

Download workflow.json and import into n8n: Workflow menu → Import from File

📖 Importing guide · 🔑 Credential setup