api-watchdog

---

About

api-watchdog is an uptime monitor written in Go It pings your http endpoints and alerts when an API goes down (and when it recovers)

Feature	Description
HTTP checks	Monitor urls on a set interval
Latency tracking	tracks response times in notifications
Status monitoring	Marks a service as down if it doesn’t return 200
Discord notifications	Posts clear DOWN and RECOVERED messages

---

Tech Stack

• Go (1.22+)

---

Setup

# Clone & enter project
git clone https://github.com/yourpov/api-watchdog
cd api-watchdog

# Install deps
go mod tidy

# Build
go build -o api-watchdog .

---

Configuration

{
  "global": {
    "discord_webhook": "",
    "request_timeout_ms": 8000
  },
  "checks": [
    { "name": "API 1", "url": "https://api.com/", "check_interval_seconds": 60 },
    { "name": "Local API", "url": "http://127.0.0.1:8080","check_interval_seconds": 15 }
  ]
}

Notes

• request_timeout_ms: how long to wait before marking a request as failed
• Each check needs a name, url, and check_interval_seconds
• All checks default to HTTP GET and expect 200 OK

---

Run

# Run once
./api-watchdog -once

# Run continuously
./api-watchdog -config ./config/config.json

---

Alert Behavior

• Sends DOWN when a healthy check becomes unhealthy
• Sends RECOVERED when an unhealthy check becomes healthy again
• Each alert includes the endpoint URL and measured latency (if available)

---

Troubleshooting

• No alerts sending
  Make sure discord_webhook is a a valid webhook

• Always shows “DOWN”
  Make sure the API returns 200 OK (you can adjust the code if you want to allow other statuses)

• False alarms from slow APIs
  Raise request_timeout_ms in global

• Multiple APIs fire at the same time
  Checks are staggered, but you can raise check_interval_seconds to send them further apart

---

Showcase

▶ Watch the showcase