Demo Service

A tiny Go HTTP service with synthetic failures that make it easy to demonstrate problem detection.

---

Features & Endpoints

Path	Purpose	Typical Error Scenarios
/	Health check / welcome JSON	—
/panic	Launches a goroutine that panics; recovered so the server lives	level=error msg="recovered goroutine panic" …
/slow	Sleeps 6 s; if the client aborts early we log context cancellation	level=error msg="context canceled" …
/migrate	Runs an intentionally broken SQL migration	level=error msg="migration failed" …
/health	Lightweight liveness probe, no extra logging	—

---

Quick Start

git clone <repo‑url>
cd demo‑service

go run .

Server starts on :8080:

level=info msg="starting server" addr=:8080

---

Playing with the Failures

Below are example requests and the exact log lines you should expect so you can wire them into your detection rules.

The duration=… value will vary, so you can replace it with .* in regexes.

/ – baseline request

curl -s http://localhost:8080/ | jq

Logs:

level=info method=GET path=/ status=200 duration=…

---

/panic – goroutine panic

curl -s http://localhost:8080/panic | jq

Logs:

level=info method=GET path=/panic status=200 duration=…
level=error msg="recovered goroutine panic" panic=intentional panic inside goroutine for demo purposes

---

/slow – client abort

Trigger with a 1‑second timeout so the client drops before the 6‑second sleep finishes:

curl --max-time 1 http://localhost:8080/slow || true

Logs:

level=info method=GET path=/slow status=200 duration=…   # emitted after handler returns (if it returns)
level=error msg="context canceled" path=/slow err=context canceled

If you let it run the full 6 s instead, you’ll just see a normal status=200 line.

---

/migrate – SQL migration failure

curl -v http://localhost:8080/migrate

Logs:

level=info method=GET path=/migrate status=500 duration=…
level=error msg="migration failed" err=alter table: no such table: imaginary

---

/health – liveness probe

curl -s http://localhost:8080/health

No extra logs (handler bypasses the logging middleware).

---

Customising

• Add new problem cases by creating a handler that logs at level=error.
• Switch log format or destination by editing log.SetFlags / log.SetOutput in main.go.

Happy hunting! 🚀