Cloud probes can't reach 10.0.0.0/8
UptimeRobot, Pingdom, and BetterStack run from the public internet — they can't ping your internal MikroTik router, office printer, NAS, or AWS VPC endpoints.
Self-host a tiny Docker probe inside your network. PingZen does the rest — multi-channel alerts, incident history, public status pages — and nothing in your network has to be opened to the internet.
UptimeRobot, Pingdom, and BetterStack run from the public internet — they can't ping your internal MikroTik router, office printer, NAS, or AWS VPC endpoints.
Self-hosted alternatives require a dedicated server, agents, templates, and ongoing maintenance. Great for ISPs and NOCs, overkill for an office LAN.
Mixing public and private monitoring across two tools means two dashboards, two alert channels, and two on-call rotations. Private Probe keeps everything in PingZen.
A Private Probe is a single Docker container you run inside your network. It opens an outbound WebSocket connection to PingZen (no inbound ports, no static IP, NAT-friendly). It gets the list of its monitors and runs the checks on its own schedule against RFC1918 hosts (10.x, 172.16-31.x, 192.168.x); if the connection drops, it keeps checking and sends up to an hour of buffered results when it reconnects. 150–400 MB of RAM depending on the number of monitors; runs on Linux/amd64, Linux/arm64, or macOS via Docker Desktop.
services:
pingzen-checker:
image: registry.gitlab.com/dobryj/pingzen/checker:0.9.27
restart: unless-stopped
cap_add: [NET_RAW]
environment:
PINGZEN_URL: https://pingzen.dev
PROBE_KEY: pz_your_api_key_here
PROBE_NAME: "Office LAN"
PROBE_TYPE: home
PROBE_REGION: "Office, RU"Create a pz_ API key (avatar at the bottom of the sidebar → API Keys), then paste it into PROBE_KEY. After docker compose up -d, the probe appears under Check Locations in the monitor form within 30 seconds.
What the Private Probe actually monitors inside a typical office or home network.
Check that the router answers inside the LAN: ping the gateway, TCP-check the API port, monitor the web UI. To catch a WAN outage, monitor the router's public IP from a public checker.
ICMP · 192.168.88.1 TCP · 192.168.88.1:8728 (RouterOS API) HTTP · http://192.168.88.1/login
Catch printer outages before users complain. Most network printers expose IPP (port 631) and a web UI for ink/toner status.
TCP · 192.168.1.50:9100 (RAW print) HTTP · http://192.168.1.50/ (printer status)
Make sure your backup target is online. Check the web admin, SMB share availability, and the DSM/QTS API.
HTTP · http://192.168.1.20:5000 (DSM) TCP · 192.168.1.20:445 (SMB) TCP · 192.168.1.20:5001 (DSM HTTPS)
The Private Probe runs every PingZen check type; below is what intranet uptime usually needs. PageSpeed needs a Google PageSpeed API key on the probe, and Transaction checks need a probe image with a browser: we don't publish one — build it with INCLUDE_BROWSER=true and give it 2 GB of RAM.
SNMP is not yet supported. If you need SNMP graphs (CPU, RAM, interface counters) on top of uptime, leave feedback in the dashboard — it's the next protocol on the consideration list.
How many Private Probes you can register on your account.
| Plan | Private probes per account |
|---|---|
| FREE | 10 (the only plan for now) |
A free account includes 10 Private Probes. The Docker container takes 60 seconds to start.