HTTP / HTTPS
Status codes, latency, redirects and response assertions on body, headers or JSON.
Self-hosted · single binary
Know the moment
Know the moment
something goes down.
AlertX is a fast, secure, self-hosted uptime monitoring and alerting platform. HTTP, TCP, Ping, DNS and SSL checks, response-time graphs, incidents, status maps and dependency-aware alerting — shipped as a single ~20 MB binary with the UI built in.
- Runs anywhere Docker or a binary does
- SQLite out of the box, Postgres when you scale
- Your data stays on your infrastructure
~20 MBSingle static binary
6Probe types built in
12Notification channels
100%REST-API covered
Monitor everything
One agent, every check you need
Point AlertX at a URL, a host, a port or a certificate. Each probe returns up/down, latency and a message — graphed, scored and ready to alert on.
TCP
Raw port reachability and connect time for any service that listens.
Ping (ICMP)
Classic host liveness and round-trip latency, loss-aware.
DNS
Resolve records and verify answers — catch poisoning and drift.
SSL / TLS
Certificate validity and expiry — warn days before a cert lapses.
Push / Heartbeat
Cron jobs and workers ping AlertX; miss the window and it alerts.
Built for operators
Everything a real monitoring stack needs
Not just checks — the incidents, context and control to actually run a service.
📈
Response-time graphs
Latency history per monitor with hourly and daily roll-ups, so long-range trends survive pruning.
🚨
Incidents & timeline
Every outage is recorded with a timeline, duration and resolution — not just a red dot that disappears.
🧩
Dependency-aware alerting
Declare what depends on what. When a root cause fails, the downstream noise is suppressed — one alert, not fifty.
🗺️
Status maps
Lay out your estate as nested, drillable maps with typed links — label a connection an IPsec tunnel, a WAN or a peering.
🌐
Public status pages
Share live availability with customers on a clean public page, without exposing your internal console.
🔔
12 notification channels
Email, Slack, Discord, Teams, Telegram and more — with escalation policies and flap suppression.
👥
Multi-user & RBAC
Admin, editor and viewer roles, TOTP two-factor and optional single sign-on via OIDC.
🔑
Single sign-on (OIDC)
Log in with Authentik, Keycloak, Entra ID, Google or Okta — with group-to-role mapping.
🗄️
SQLite or Postgres
Zero-config SQLite for a homelab; switch to PostgreSQL when you need scale and HA — same binary.
Status maps
See your estate, not just a list
Group monitors into buildings, sites and regions, then nest them as deep as your topology goes. Draw the links between them and label each one — an IPsec tunnel, a leased line, a peering — so a glance tells you how the pieces connect and where it hurts.
- Typed, labelled links between any two nodes
- Drill from region → site → rack → monitor
- Shareable on public status pages
Signal, not noise
Alert on the cause, not the symptoms
When a switch, a link or a database goes down, everything behind it looks down too. AlertX understands those dependencies: it still records every incident, but it suppresses the downstream alerts and pages you about the one thing that actually broke.
- Transitive dependency graph (A → B → C)
- Downstream outages recorded but kept quiet
- Escalation policies and flap detection
Editions
Start lean. Grow when you need to.
Same binary, same UI. Plus unlocks distributed monitoring for multi-site estates.
Local
The complete single-node platform.
- All six probe types
- Incidents, graphs & status pages
- Status maps & dependency alerting
- 12 notification channels
- Multi-user, RBAC, OIDC & TOTP
- Full REST API & Prometheus metrics
Plus
Plus
Everything in Local, for distributed estates.
- Everything in Local
- Remote probe agents — check from many locations
- Per-monitor probe location
- Subnet discovery (CIDR sweeps)
- Agent-run checks across sites
Notifications
Reach your team where they already are
- Slack
- Discord
- Microsoft Teams
- Telegram
- Webhook
- ntfy
- Gotify
- Pushover
- Pushbullet
- Matrix
- Custom REST
For developers
Everything the UI does, the API does
AlertX ships a complete REST API and a Prometheus /metrics endpoint, so monitors are code, dashboards are yours, and automation is first-class — not an afterthought.
- Full REST API with scoped API keys
- Prometheus-compatible metrics export
- JSON backup & restore of your config
# create a monitor from the command line
curl -X POST https://alertx.example.com/api/v1/monitors \
-H "Authorization: Bearer $ALERTX_KEY" \
-H "Content-Type: application/json" \
-d '{"name":"api","type":"http",
"target":"https://api.internal/health",
"interval_seconds":30}'
{ "id": 42, "status": "pending", "ok": null }
Self-hosted by design
Your monitoring, your machine, your data
No SaaS account, no per-monitor pricing, no telemetry phoning home. Deploy it with one container and own every byte.
One container
A single image with the UI embedded. docker compose up and you are monitoring.
No lock-in
SQLite or Postgres, JSON backups, an open REST API. Take your data anywhere.
Private
Runs entirely inside your network. Nothing leaves your infrastructure.
Deployment
Run it yourself — or let us run it
Self-hosting is the default and always will be. When you'd rather not own the ops, we'll host and operate a dedicated instance for you.
No VPN. No open ports. Remote sites run a single agent that dials out to the hub over HTTPS — so you can monitor branch offices and customer locations without inbound firewall rules, port-forwarding or a mesh VPN. It works behind any NAT.
Self-hosted
Your infrastructure, your data, your control.
- Runs entirely inside your network
- One container or binary; SQLite or Postgres
- Remote sites via outbound-only agents
- You own upgrades and backups
Hosted by us
A dedicated instance we run, so you don't have to.
- We operate, patch, upgrade and back it up
- A dedicated, isolated instance — not shared
- Your sites connect with the same outbound agents
- Ideal for multi-site estates without the ops overhead
Want to see AlertX in action?
Get in touch for a walkthrough, early access or just to talk monitoring.