Telegraws

AWS Infrastructure Monitoring with Telegram Notifications

Telegraws is a lightweight, serverless AWS monitoring scheduled Lambda function that collects metrics from multiple AWS services and sends automated reports to Telegram.

Features

• Serverless: Deploys as AWS Lambda function with automatic scheduling.
• Telegram Integration: Sends formatted monitoring reports to Telegram.
• IAC: Automatically creates IAM roles, Lambda functions, and EventBridge schedules.
• Local Development: Test locally with --local flag before deployment.
• Multi-Service Monitoring: EC2, S3, ALB, CloudFront, DynamoDB, RDS, WAF, CloudWatch Logs, Cloudwatch Agents.
• Smart Scheduling: Hourly updates + daily reports.
• Immutable Deployments: Clean, reproducible deployments.

Prerequisites

• AWS CLI installed and configured in the same region as your monitoring infrastructure.
• Go 1.24+ installed.
• Telegram Bot token and chat ID.
• AWS Resources to monitor.

Usage

git clone https://github.com/victorvibes/telegraws.git
cd telegraws
go mod tidy
cp config/config-template.json config/config.json
# Edit config/config.json with your settings
./build.sh --lambda # or --local

Considerations

• Running ./build.sh --lambda automatically detects if the function was already deployed and updates the existing function.
• Telegraws creates these resources: Lambda function, IAM role, Eventbridge rule. All prefixed by 'telegraws-'.
• lambdaCronExpression: EventBridge cron schedule (AWS format: Minutes Hours Day Month DayOfWeek Year). Requires raw expression, eg: "lambdaCronExpression": "0 * * * ? *".
• timezone: Go time.LoadLocation compatible timezone.
• defaultPeriod: Hours to look back for regular reports (1 = last hour). Set to 0 to only receive daily reports.
• dailyReportHour: Hour to send daily summary (respects timezone).
• CloudWatch Logs collection counts INFO/WARN/ERROR so structured logging is required.
• RDS monitoring currently supports Aurora engine.
• WAF monitoring collects WAFs metrics attached to ALB.
• CloudWatch Agent monitors disk_used_percent and mem_used_percent.
• Telegram has 4096 character limit per message.

Metrics

• EC2: CPU Utilization (avg/max), Network I/O, Status Checks. If CloudWatch Agent: mem_used_percent, disk_used_percent.

• S3: (Daily Reports Only) Bucket Size, Objects Count.

• ALB: Request Count, Response Time, HTTP Status Codes, Healthy/Unhealthy Hosts, ALB Errors.

• CloudFront: Requests, Bytes Uploaded, Bytes Downloaded, Error Rates.

• DynamoDB: Request Count, Items Count, Throttles, Latency, Consumed Capacity, Error Counts.

• RDS/Aurora: Instance: CPU, Memory, Connections, Read/Write Latency. Cluster: Volume Size, IOPS.

• WAF: Allowed/Blocked Requests.

• CloudWatch Logs: INFO/WARN/ERROR log counts (requires structured logging).

To-do

• Enhanced Metrics: Add comprehensive metric collection for all services. Get metrics dynamically using AWS CLI?
• Dynamic Metrics: User-configurable metrics selection. Separated daily and scheduled metrics.
• Enhanced AWS Support: ECS/EKS, Fargate, API Gateway.
• Multi-Resource: Multiple IDs per service type.
• Message Splitting: Handle Telegram 4096 character limit.
• Cross-Platform: Windows support for build script.
• Emoji Support: Optional emoji integration in messages.
• Architecture Options: x86_64 Lambda support.
• RDS Engines: Support for MySQL, PostgreSQL, SQL Server.
• Advanced WAF: Multiple WAF configurations.