Lynx · Server & container monitoring

Your servers learn to watch themselves.

A lightweight agent on every server and container. It learns what normal looks like and alerts you only when something is actually wrong.

Free up to 5 nodes. Linux and Docker.

# Install on any Linux server
$ curl -sL https://luviner.com/install-lynx.sh | bash

# Done. Monitoring is live.
✓ Agent installed (3.4 MB)
✓ Learning baseline... done (5 min)
✓ Monitoring active. Dashboard: https://luviner.com
§01The problem

Static thresholds cry wolf.

CPU at 95% on Friday night is the backup, not an emergency. Lynx knows that.

SituationStatic thresholdsLynx
Who decides what is normalYou, by hand, server by serverLynx, on its own
Friday backup, CPU at 95%AlertSilence: that's normal
Slow memory leakYou find out at the crashAlert days before
Deploy: 5 servers spike together5 alertsNo alert: the nodes recognise the deploy
§02What it does

A brain on every node.

01

Anomalies

Each server learns its own normal. It adapts on its own to deploys, seasonality and new machines.

02

Drift

Memory leaks, filling disks, slowing performance: caught days before the crash.

03

Mesh consensus

Nodes vote. Five servers spiking together: a deploy. One alone: a problem.

04

Light and private

Under 5 MB of RAM, under 0.1% CPU. Raw metrics stay on your server.

§03Docker

Finds every container by itself.

One model per container. New containers are picked up within 30 seconds.

# Install with Docker monitoring
$ curl -sL https://luviner.com/install-lynx.sh | bash -s -- --docker

# Or run Lynx itself as a container
$ docker run -d \
  -v /var/run/docker.sock:/var/run/docker.sock \
  -e LYNX_DASHBOARD_URL=https://luviner.com \
  -e LYNX_API_TOKEN=your_token \
  luviner/lynx:0.1.0 --docker
§04Setup

Running in 5 minutes.

  1. Install

    One command. The agent installs and registers itself.

  2. Learn

    It watches the server for 5 minutes and learns what normal is.

  3. Detect

    Continuous monitoring, real-time alerts.

  4. Decide together

    With several servers, nodes tell a deploy from a real fault.

60s
Install to first data
12
System metrics per node
0
Thresholds to set
100%
On-device: data never leaves

Native on Linux: Ubuntu, Debian, RHEL, x86_64 and aarch64. On macOS the same command runs Lynx in Docker (Docker Desktop required).

§05The engine

The same engine as Shield.

Lynx runs on Vivi, our proprietary AI engine.

  1. 01 Shield Website protection Stops bots, scrapers and traffic waves. Customers and Google get through. →
§06Pricing

One flat price per node tier.

Starter
Free

Up to 5 nodes

Team
€79/mo

Up to 25 monitored nodes

Scale
€499/mo

Up to 500 monitored nodes

Enterprise Beyond 500 nodes or regulated environments: air-gapped deployment, dedicated support, custom SLA. Get in touch
Full plan details →

No more false alarms.

Free up to 5 nodes. One-command install.

Start free → See pricing