Uptime Monitor

Uptime Monitor is a serverless AWS project built with Go, Lambda, EventBridge, S3, OpenTofu (Terraform-compatible), and GitHub Actions.

It monitors multiple websites on an hourly schedule, stores the latest status snapshot and recent history in S3, and exposes lightweight HTTP endpoints through a Lambda Function URL for manual checks and dashboard data.

Full Documentation

---

Case Studies

Case Study	Problem	How it was diagnosed	Result
Why S3 JSON storage	The monitor needed persistent latest status and short per-site history without adding database infrastructure.	S3 fit because the MVP only needed small structured writes, current-state reads, and short history reads; it did not need relational queries, indexes, joins, or high-write throughput.	Results are stored as latest.json and history.json, keeping storage low-cost, frontend-friendly, and database-free for the MVP.
Why EventBridge scheduling	The monitor needed to run automatically without relying on a user or external client to call /check.	EventBridge fit because it keeps the scheduler serverless, avoids a VM cron job or always-on worker, and supports the hourly cadence needed for the first version.	EventBridge invokes the Lambda every hour with rate(1 hour), reusing the same monitor path as manual checks.

---

Architecture

The system runs as a small serverless monitoring loop: Terraform provisions AWS resources, EventBridge triggers scheduled Lambda checks, and S3 stores the latest and historical status data.

Path	Use case	Flow
Scheduled monitoring	Check configured websites every hour	EventBridge -> Lambda -> monitored sites -> S3
Manual operation	Trigger checks or read monitor data on demand	Lambda Function URL -> /check, /latest, /history, /health
Runtime storage	Keep latest and recent historical status data	Lambda -> S3 latest.json and history.json
Infrastructure deployment	Keep AWS resources managed from source control	GitHub Actions -> Terraform -> AWS resources

Deployment flow:

flowchart LR
    Source["GitHub Repository"]
    Actions["GitHub Actions"]
    Terraform["Terraform"]
    AWS["Lambda, S3, EventBridge, IAM"]

    Source --> Actions
    Actions --> Terraform
    Terraform --> AWS

Loading

Runtime flow:

flowchart LR
    Schedule["EventBridge"]
    Lambda["AWS Lambda"]
    Health["GET /health"]
    Check["POST /check"]
    Latest["GET /latest"]
    History["GET /history"]
    Sites["Monitored sites"]
    Store["Storage layer"]
    LatestFile["latest.json"]
    HistoryFile["history.json"]
    S3["S3 bucket"]

    Schedule -- every hour --> Lambda
    Lambda --> Health
    Lambda --> Check
    Lambda --> Latest
    Lambda --> History
    Check -- run checks --> Sites
    Sites -- results --> Check
    Check -- save results --> Store
    Store -- write --> LatestFile
    Store -- update --> HistoryFile
    LatestFile -- store --> S3
    HistoryFile -- store --> S3
    Latest -- read latest.json --> S3
    History -- read history.json --> S3

Loading

---

Tech Stack

Layer	Tools
Language	Go
Compute	AWS Lambda
Scheduling	Amazon EventBridge
Storage	Amazon S3
Infrastructure	Terraform
Testing	Go testing package, table-driven tests
CI/CD	GitHub Actions

---

API Endpoints

Endpoint	Method	Purpose
/health	GET	Returns Lambda service health
/check	POST	Runs uptime checks and writes results to S3
/latest	GET	Returns the latest status snapshot
/history	GET	Returns recent status history grouped by URL

---

Documentation

• Documentation Hub
• Architecture
• Decisions
• Incidents

---

Local Setup

Build and test the Go backend:

make test
make lambda-package

Deploy infrastructure locally:

cd infra
tofu init
tofu apply

Trigger a manual check:

curl -X POST "$(tofu output -raw lambda_function_url)check"

Read monitor data:

curl "$(tofu output -raw lambda_function_url)latest"
curl "$(tofu output -raw lambda_function_url)history"