Developers
Monitoring you can add without ceremony.
Website, API, SSL, domain and DNS checks need nothing installed. Server metrics need one small agent, and you can put it on a machine in a single command.
Installing the server agent
Add the server in the dashboard and it gives you a command with that server's own install URL in it. The script detects the operating system and architecture, fetches the matching binary and registers it as a service, so it comes back after a reboot.
- Linux, Windows, macOS and FreeBSD on amd64 or arm64
- Needs root, because it writes to /usr/local/bin and /etc/systemd
- Outbound HTTPS only, no inbound port opened
- Or hand PingXen SSH credentials and let it install the agent for you
# Linux and macOS: the dashboard fills in the server id for you
curl -sSL https://pingxen.com/agents/SERVER_ID/install-script/linux/amd64 \
| sudo bash
# Windows, from an elevated PowerShell
powershell -ExecutionPolicy Bypass -Command "Invoke-WebRequest
-Uri '.../install-script/windows/amd64'
-OutFile install.ps1; .\install.ps1"What the agent sends, and how often
It collects on an interval you set and posts to one endpoint. Nothing else leaves the machine.
The metrics
Hostname, operating system, CPU, memory, disk usage per mount and load average. No file contents, no process arguments, no logs.
The interval
Five minutes by default. Paid plans can go down to one minute; the free plan is capped at ten.
When the network drops
The agent keeps running and keeps retrying, and it starts again with the machine, so an outage on your side does not mean reinstalling it.
Checks that need nothing installed
Everything except server metrics runs from our side against an address you give us, so you can monitor infrastructure you do not have shell access to.
- HTTP and HTTPS with your own method, headers, body and basic auth
- Keyword rules for text that must, or must never, be in the response
- Certificate expiry, chain and hostname validation
- Domain expiry, resolved through RDAP with a WHOIS fallback
- A, AAAA, CNAME, MX, NS and TXT record comparison
Honest note on the API
PingXen has API keys, but today they authenticate the server agent rather than a public management API. There is no documented endpoint for creating monitors from your own code yet.
If provisioning monitors from a deploy pipeline is something you need, say so. It is on the list, and knowing who wants it moves it up.
Create your first monitor
Start with a URL. Add the agent later, when you want the server behind it in view too.