What is LAN agent and when to use | Help & guides ePulz.io

In short: A LAN agent is a small program that you install on any computer inside the client's internal network. It connects over the internet to ePulz.io, requests the list of addresses it should check, verifies them locally, and sends the results back. Thanks to this, you can monitor even devices that cloud monitoring cannot reach (NAS, cameras, internal servers, local web applications).

When you need a LAN agent

Monitoring NAS and home or business servers - Synology, QNAP, your own machine with ZFS storage
IP cameras, access control systems, alarms in a shop or office
Internal web applications - Grafana, Jira, GitLab, Bookstack behind a firewall
Network infrastructure - router, switches, access points (AP), VPN gateway
IoT devices and gateways - Modbus and MQTT gateways and the like
MSP scenario - managing monitoring for multiple clients with a single, unified deployment

How it works

In the dashboard you create an agent and receive a token (a unique key) epulzio_… - it is shown only once.
You run the install command on a device in the client's network (Raspberry Pi, mini PC, NAS, a computer with Docker).
Every 30 seconds the agent asks ePulz.io for new tasks and runs the local checks.
It sends the results back over HTTPS - the same secure connection that every browser uses.
When a device's status changes, ePulz.io triggers the standard alert (e-mail, Telegram, webhook).

Important security note: All communication only goes out of your network (outbound HTTPS on port 443). You never need port forwarding on the router, a static public IP, or a VPN tunnel. From the perspective of the client company's IT security, this is just as commonly acceptable as opening Gmail.

What the agent sends (and what it never sends)

The agent sends exactly three things:

Heartbeat - the agent version, hostname, and operating system name
Check results that you configured yourself - up/down status, response time in ms, and any error message
Nothing else. It never scans the network, reads files, or collects traffic or device data. Every single IP address it touches is defined by you in the dashboard.

Limited token: the agent token only allows it to download its own list of checks and submit results. It has no access to your account, billing, or other monitors, and you can revoke it at any time in the dashboard (Revoke).
Runs without root privileges: the service runs as an unprivileged system user with a systemd sandbox (NoNewPrivileges, ProtectSystem=strict). Root is only needed during installation.
Fully auditable: the entire agent is a single Python file (~600 lines, standard library only) - before running it you can read every line.

Three types of LAN checks

Type	Target	What it detects
lan_ping	192.168.1.10	The device responds to ICMP (a simple network test). The cheapest check.
lan_tcp	10.0.0.50:5432	A specific port is open. Even if the device is running, the service itself (database, SSH, and the like) may be down.
lan_http	http://nas.local	The HTTP or HTTPS application responds correctly (2xx/3xx codes). Suitable for a NAS interface or an internal dashboard.

List of LAN agents in the ePulz.io dashboard — List of LAN agents - status, last contact, and version of each agent

What happens when an agent stops working

Every minute ePulz.io checks whether the agents are sending a sign of life (heartbeat). If an agent does not report for more than 5 minutes, you receive both an e-mail and a Telegram message: "LAN agent X is offline". When the agent returns, you get a confirmation that it is back online. This way you avoid the false sense that everything is running, just because an agent stopped sending data.

Get started

In the next step we will show you how to install the agent and create your first LAN monitor.

Next step:

Installing the LAN agent on a device in the client's network.

Install the agent ->