Monitor Types
Pick the monitor that matches the evidence you can collect. If the target can call KEEPitALIVE, use Heartbeat/Event Receiver. If KEEPitALIVE must call the target, use HTTP/API, Browser, Ping, TCP, DNS, or Game.
CHOOSING FAST
- Public HTTP health endpoint? Use HTTP.
- JSON API response needs field assertions or captured response details? Use API.
- Web app only works after JavaScript, login-wall detection, or WAF/bot behavior matters? Use Browser.
- Host reachability only? Use Ping.
- Port must accept connections? Use TCP.
- DNS answer must match expected records? Use DNS.
- Cron job, backup, deploy, worker, or private network signal? Use Heartbeat.
- Minecraft, FiveM, or Steam/Source server? Use Game Server.
HTTP
Polls a URL and treats selected status codes as healthy. Use it for simple health endpoints, public APIs, landing pages, redirects, TLS/domain-expiry checks, keyword present/absent checks, and latency thresholds.
GET https://api.example.com/healthAPI
HTTP monitoring with request bodies, content type, custom headers, expected statuses, JSON assertions, captured response body/headers, and API-specific diagnostics. Use it when "200 OK" is not enough.
POST https://api.example.com/graphql + JSON assertionBrowser
Runs a headless browser against the page. Use it when JavaScript rendering, SPA boot, client-side errors, or bot/WAF behavior matters. Browser checks are heavier and cost more credits.
browser https://app.example.comPing (ICMP)
Sends ICMP ping and measures round-trip time. Good for network appliances, hosts, and public IPs where "reachable" is the signal. Not useful when ICMP is blocked by design.
ping monitor.example.netTCP Port
Attempts a TCP connection to host:port. Use it for SSH, SMTP, MQTT, databases, queues, proxies, and any service where accepting a socket is enough proof of availability.
tcp://db.internal:5432DNS
Resolves one or more DNS record types and optionally asserts expected values. Use it for A/AAAA/CNAME/MX/TXT/NS drift detection, resolver comparison, and domain-expiry tracking.
A + TXT records for example.comHeartbeat / Event Receiver
Inverted monitoring. Your job, deploy pipeline, agent, or private network gateway POSTs to KEEPitALIVE. Scheduled mode alerts on missing beats; Event-Receiver mode evaluates payload conditions.
POST heartbeat_url with {"status":"maintenance"}Game Server
Queries Steam/Source A2S, Minecraft, FiveM, or TCP/ping fallbacks. Tracks availability, latency, map/server metadata, player count, and optional player roster where supported.
A2S_INFO udp://game.server:27015WHAT NOT TO MONITOR DIRECTLY
- Do not point checks at sensitive internal services or control panels unless they are intended to be monitored from KEEPitALIVE.
- Do not use Browser when HTTP/API gives the same signal; it is more expensive and slower.
- Do not use Ping/TCP when you need semantic correctness. A port can be open while the application is broken.
- Do not use Event Receiver as a recovery detector unless your sender always posts the closing payload. Silence is not treated as recovered.