LNX Monitor

UniFi Site Manager and network monitoring for managed service providers.

LNX Monitor is the network operations and monitoring platform built by LNX Inc. for managed service providers. It syncs sites and devices from UniFi Site Manager, raises incidents when a site goes offline or degrades, and tracks infrastructure health across servers, websites, SSL certificates and third-party telemetry delivered by webhook. Checks run on a schedule through dedicated drivers (HTTP, ICMP ping, TCP port, SSL certificate, SIP) with alert routing, maintenance windows and public status pages built in.

Built with Laravel 13, Inertia 2 and Vue 3.

What it does

• Multi-protocol monitoring with a driver architecture: HTTP, ping, TCP, SSL, SIP and webhook ingest
• Per-monitor policies and notification channels: email, Slack webhooks, generic webhooks
• Queued notification delivery with automatic retries, de-duplication and flood control
• Maintenance windows to suppress noise during planned work
• UniFi Site Manager integration for automated site and device sync
• Public-facing status pages
• Centralised incident tracking with resolution notes

Quick start with Docker

The published image contains the app, compiled front-end assets and every PHP extension it needs. The compose file adds MySQL and Redis and runs the web server, a queue worker and the scheduler from that one image. Nothing has to be built or configured on the host.

mkdir lnx-monitor && cd lnx-monitor
curl -fsSLO https://raw.githubusercontent.com/lnxinc/monitor/master/compose.yaml
docker compose up -d

Open http://localhost:8080/register and create the first user. That's it.

On first start the app container generates an application key and stores it in the storage volume, runs the database migrations, links public storage and caches configuration. Later starts reuse the same key and run only new migrations.

The stack

Service	Image	Role
app	ghcr.io/lnxinc/monitor	NGINX + PHP-FPM serving the UI and API on port 8080
queue	ghcr.io/lnxinc/monitor	queue:work runs monitor checks and sends notifications
scheduler	ghcr.io/lnxinc/monitor	schedule:work dispatches due monitors and UniFi sync every minute
mysql	mysql:8.4	Application database, persisted in the mysql volume
redis	redis:7-alpine	Cache, sessions and queue, persisted in the redis volume

Every container reports health, and the queue and scheduler wait for the web app to pass its health check before starting, so migrations always finish before jobs run.

Configuration

Every setting has a working default. To change one, create a .env file next to compose.yaml (start from .env.docker.example) and run docker compose up -d again.

Variable	Default	Purpose
MONITOR_VERSION	latest	Image tag to run. Pin a release such as 1.2.0 in production.
APP_PORT	8080	Host port the UI is published on.
APP_URL	http://localhost:8080	Public URL, used in emails and generated links.
APP_KEY	generated	Set to reuse an existing key. Leave blank to auto-generate once.
DB_DATABASE / DB_USERNAME / DB_PASSWORD	monitor	MySQL credentials. Change them before exposing the stack.
DB_ROOT_PASSWORD	monitor-root	MySQL root password.
MAIL_*	log mailer	SMTP settings for outbound alerts.
UNIFI_SITE_MANAGER_API_KEY	empty	Enables the UniFi Site Manager sync.
HEARTBEAT_*, ALERT_*	see below	Reliability tuning.

Behind a reverse proxy or load balancer, set APP_URL to the public address and terminate TLS at the proxy. The app listens for plain HTTP on 8080.

Day-to-day operations

docker compose pull && docker compose up -d      # update to the newest image
docker compose logs -f app queue scheduler        # follow application logs
docker compose exec app php artisan monitors:run  # run every due monitor now
docker compose exec app php artisan tinker        # poke at the app
docker compose exec mysql mysqldump -umonitor -pmonitor monitor > backup.sql

Three volumes hold all state: storage (application key, uploads, logs), mysql and redis. Back those up and the stack can be recreated anywhere with the same compose file.

Building the image yourself

git clone https://github.com/lnxinc/monitor.git && cd monitor
docker compose -f compose.yaml -f compose.dev.yaml up -d --build

Or build the image on its own with docker build -t lnx-monitor:local .. The Dockerfile is a two-stage build: stage one installs Composer and npm dependencies and compiles the Vite bundle, stage two copies the result into an unprivileged PHP-FPM + NGINX runtime. Pushes to master and v* tags publish multi-architecture images (amd64 and arm64) to GHCR through the docker workflow.

Running without Docker

Requirements

• PHP 8.4 with bcmath, ctype, curl, mbstring, openssl, pdo and a database driver
• Composer 2
• Node.js 22 and npm
• MySQL 8, MariaDB 10.5, PostgreSQL 13 or SQLite
• Redis for cache, sessions and queue (the database driver works too)
• A process manager for the queue worker and cron for the scheduler

Setup

git clone https://github.com/lnxinc/monitor.git && cd monitor
composer install --no-interaction --prefer-dist
npm install
cp .env.example .env
php artisan key:generate
php artisan migrate --force
npm run build

Local development

composer run dev starts the web server, queue listener, log tail and Vite together. Run them separately if you prefer:

php artisan serve
php artisan queue:listen --tries=1
php artisan schedule:work
npm run dev

Production deployment

1. composer install --no-dev --optimize-autoloader
2. npm ci && npm run build
3. Set production values in .env
4. php artisan migrate --force
5. php artisan config:cache && php artisan route:cache && php artisan view:cache
6. Supervise php artisan queue:work --sleep=1 --tries=3
7. Cron: * * * * * php /path/to/artisan schedule:run >> /dev/null 2>&1
8. Give the web server user write access to storage/ and bootstrap/cache/
9. On Linux, the ping driver needs the PHP process to run ping. Grant it with setcap cap_net_raw+ep $(command -v ping).

Monitoring and maintenance

• Run every due monitor by hand: php artisan monitors:run, or use Run now on a monitor in the UI.
• Webhook ingestion: POST /api/webhooks/monitors/{token} with a JSON payload.
• Health endpoint for load balancers: GET /up.
• Logs go to stderr in Docker (docker compose logs) and to storage/logs/laravel.log otherwise.

Reliability tuning

Variable	Default	Effect
HEARTBEAT_MAX_ATTEMPTS	3	Retries for transient check failures
HEARTBEAT_INITIAL_BACKOFF_MS	200	Delay before the first retry
HEARTBEAT_BACKOFF_MULTIPLIER	2.0	Multiplier applied to each following retry
ALERT_DEDUPE_TTL_SECONDS	300	Window for suppressing identical incident notifications
ALERT_FLOOD_LIMIT	10	Messages a channel may receive per flood window
ALERT_FLOOD_WINDOW_SECONDS	60	Flood-control window

Testing

php artisan test --compact
vendor/bin/pint --dirty
npm run lint && npm run format:check

CI runs the Pest suite, Pint and the front-end linters on every push and pull request.

Brand

LNX Monitor is an LNX Inc. product. Anything this project puts in front of a client (status pages, emails, exported reports, documentation) follows the LNX brand standard v1.0. The files live in public/brand/.

Logo

File	Use
lnx-logo-primary.png	Primary lockup, light backgrounds
lnx-logo-reversed-white.png	White lockup, transparent. LNX Navy and Deep Navy backgrounds only.
lnx-logo-mono-navy.png	One colour, single-ink print
lnx-mark-gradient.png	Mark alone, for avatars, favicons and badges
lnx-gradient-rule.png	The gradient rule as an image, for engines that cannot render CSS gradients

Lockups are 2000 × 646, a 3.096:1 ratio. If the width divided by the height is not 3.096, it has been stretched. Clear space on every side equals the height of the mark. Minimum width is 120 px on screen and 1 in in print.

Never recolour, stretch, rotate, outline or shadow the logo, never place it on a busy photo, and never put the reversed mark on anything but Navy or Deep Navy. Use the supplied files; do not rebuild the mark in CSS, SVG or type.

Colour

Navy is the brand. Cyan and violet are accents for rules, keylines and a single highlighted figure, never large fills and never behind type. One accented element per page.

	Name	Hex	Role
	LNX Navy	#112540	Primary ink, headings, logo
	Deep Navy	#002657	Covers, slide and banner backgrounds
	Signal Cyan	#43BDE9	Accent, charts, rules
	Circuit Violet	#6F52F1	Accent, second data series
	Core Magenta	#B624FA	Logo core only. Never in UI, charts or highlights.
	Mist	#EFF3F8	Fills
	Line	#D3DBE5	Rules, borders
	Slate	#5A6A80	Secondary text, labels
	Body	#3D4C61	Body copy

The gradient is linear-gradient(90deg, #43BDE9 0%, #6F52F1 100%), cyan to violet, left to right. It belongs to the logo. Elsewhere it is allowed only as a rule, a keyline or a thin band at a page edge. Never behind text, never as a background, never reversed or angled.

Text on colour is limited to four combinations: white on Navy, white on Deep Navy, Navy on Cyan, white on Violet. White on Cyan fails contrast and is never used.

Typography

Two typefaces, nothing else.

• Space Grotesk (weights 500 and 600) for headlines, section titles and numbers that should be noticed. Tracking −2% at 20 pt and above. Fallback: Arial Bold.
• IBM Plex Sans (weights 400, 500, 600) for body copy, tables, captions and labels. Never below 9 pt in print. Fallback: Calibri, then Arial.
• IBM Plex Mono for codes, part numbers, specs and identifiers.

Headings are sentence case, never italic, one weight per level. Underline is for links only.

Company

Legal name	LNX Inc. (with the period)
Headquarters	1395 Wheaton Dr Suite 500, Troy, MI 48083
Phone	833-569-5690
Website	lnxinc.com

The product UI itself follows its own design-token set in resources/css/app.css. The brand standard governs what the app sends out and how the project presents itself.

License

LNX Monitor is released under the LNX Monitor Source Available License.

You are free to use, modify and deploy this software as long as you keep attribution to LNX Inc. Reselling or commercially redistributing the software as a product or SaaS offering requires prior written permission. Contact sari@lnxinc.com for commercial redistribution licensing.