Know a server is in trouble before your users do.

SRVCheck watches your Linux fleet from one place: CPU, iowait and temperature, RAM and swap, every disk partition, every network interface, and mdadm RAID health. A small Go agent pushes metrics over HTTPS every 30 seconds — nothing to poll, no ports to open.

One static binary, no dependencies · amd64 and arm64 · Any systemd Linux

Install on a server
curl -sSL https://srvcheck.com/install.php?token=YOUR_TOKEN | sudo bash

The command installs the agent binary, its config in /etc/sysmon-agent/ and a systemd unit. Metrics show up on your dashboard within a minute — you get the exact command, token included, the moment you add a server.

What it watches

Everything below is collected by default. There is nothing to configure per metric.

CPU and iowait

Usage and iowait every 30 seconds, with load averages alongside and 1h to 30d graphs that keep short spikes visible.

CPU temperature

Package temperature read straight from the kernel's hwmon sysfs — no lm-sensors, no IPMI. Virtual machines simply report none.

RAM and swap

Used, cached and available memory, plus swap pressure on the same graph so a leak is obvious before the OOM killer finds it.

Disk per partition

Every mount point tracked on its own, so a filling /var never hides behind a roomy /. Device names included.

Network per interface

rx/tx for each physical NIC with its negotiated link speed, and exact traffic totals for whatever range you are looking at.

mdadm RAID

Array state watched continuously. A degraded or failed array alerts on the next report — no waiting out an averaging window.

Uptime and reboots

Uptime per host, and a distinct alert when a machine reboots — deliberately separate from "went offline".

Threshold alerts

Per-resource rules with an averaging window and a repeat interval, set once for the whole organization or overridden per server.

How it works

1

Sign up

Create an account and name your organization. Invite the rest of the team later — everyone shares the same servers, alert rules and history, with roles from owner down to read-only viewer.

2

Run one command

Add a server, copy its curl … | sudo bash line and run it as root on the box. The agent authenticates with a per-server bearer token you can regenerate at any time.

3

Watch, and get told

Graphs from one hour to thirty days, a status dot per host, and threshold alerts to email, Discord, Pushover or your own webhook. Mute a host with maintenance mode while you work on it.

Simple plans

Every plan monitors every metric. What changes is scale, history and channels.

Compare everything

Free

Start here
Free

Kick the tyres on a handful of boxes.

  • 25 servers
  • 1 team members
  • 90 days of history
  • Reports as often as every 1 min
  • email, discord, pushover, webhook
Signups are closed

Ready to see your fleet?

Add your first server and have graphs inside five minutes.

Ask for an invitation