I found out my Jellyfin server had been down for six hours when my roommate asked why movies weren’t loading. Six hours. Of downtime I knew nothing about. That was the day I set up my uptime kuma homelab setup — and now I get a Telegram ping within 60 seconds of anything going sideways. A good uptime kuma homelab setup answers one question fast: is it down, or is it just me?
Uptime Kuma is a self-hosted monitoring tool that watches your services and yells at you when they die. It runs in a single Docker container, has a polished web UI, supports 90+ notification channels, and lets you build public status pages your family can check instead of texting you. It’s free, it’s lightweight, and it should be one of the first things you spin up in any homelab.
| Monitor Type | What It Catches | Example Use |
|---|---|---|
| TCP port checks | A service listening on a specific port going dark | Plex (32400), SSH (22) |
| Docker container heartbeats | A container stopped, even if its port still looks open | Push monitors containers ping on a schedule |
| SSL certificate expiry | A cert about to lapse | Alerts 30 days before expiry |
| DNS record checks | A domain resolving to the wrong (or no) IP | Confirming DNS changes propagated |
| Ping monitors | A host being completely unreachable | Router, NAS — if these go dark, everything else probably is too |
What Makes an Uptime Kuma Homelab Setup Worth It
Before Uptime Kuma, my monitoring strategy was “notice when something breaks.” That’s not a strategy. It’s a stress position. Uptime Kuma checks your services on a configurable interval (20 seconds by default) and blasts a notification the moment something goes offline — before anyone else in the house notices.
Here’s what you get out of the box: HTTP/HTTPS monitoring, TCP port checks, ping, DNS record validation, Docker container health, SSL certificate expiry tracking, and push-based heartbeats. Version 2.1 (released February 2026) added Globalping support, so you can check if your site is reachable from multiple global probes — not just from inside your own network. That distinction matters if you’re running services your family accesses from outside the house.
The killer feature for homelabbers is the public status page. You create a page, drag your monitors onto it, and share the URL. Now instead of “is the VPN thing down again?”, your household gets a real status page with uptime percentages and incident history. It’s genuinely satisfying to use.
Tip 1 — Run It on a Dedicated Pi (Not Your Main Server)
The whole point of monitoring is knowing when your server goes down. If Uptime Kuma runs on that server, it also goes down when the server does. Run it on something separate.
A Raspberry Pi 5 (8GB) is the ideal host for a lightweight uptime kuma homelab setup. It uses about 5–8W at idle, boots from a USB SSD in seconds, and runs Docker without breaking a sweat. Uptime Kuma itself uses under 400MB of RAM for most homelab configurations, so the Pi handles it easily with plenty of headroom for other small services.
If you’ve already got a Pi running other services, that’s okay — just be aware that if it reboots, your monitors go dark temporarily. For truly critical monitoring, a separate low-power device is the right call.
Raspberry Pi 5
If you’ve already got a Pi running other services, that’s okay — just be aware that if it reboots, your monitors go dark temporarily.
- If you’ve already got a Pi running other services, that’s okay
- just be aware that if it reboots, your monitors go dark temporarily. For truly critical monitoring, a separate low-power device is the right call
Current price and availability shown on Amazon.
Tip 2 — Plug It Into a UPS
Here’s the thing nobody tells you when you build a homelab: power outages are the number one cause of unexpected downtime, and they happen when you’re not home to deal with them. A monitoring box that goes offline during a power blip can’t tell you your router is also offline. You’re flying blind at exactly the wrong moment.
An APC BE600M1 UPS gives your Raspberry Pi and router 20–30 minutes of battery backup during a power outage. That’s usually enough to outlast brief outages and let you safely shut down if needed. It’s 600VA/330W with 7 outlets and a USB port for real-time power status monitoring from the host. It’s one of the highest-ROI purchases in any homelab setup.
Plug your Pi, your router, and your switch into the UPS battery outlets. Plug your NAS or other high-draw devices into the surge-only outlets. Now your monitoring survives most outages — and sends you a notification about the power event rather than going silent.
APC BE600M1 UPS
Plug your Pi, your router, and your switch into the UPS battery outlets.
- Plug your Pi, your router, and your switch into the UPS battery outlets. Plug your NAS or other high-draw devices into the surge-only outlets. Now your monitoring survives most outages
- and sends you a notification about the power event rather than going silent
Current price and availability shown on Amazon.
Tip 3 — Deploy It in One Command
The uptime kuma homelab setup couldn’t be simpler to actually install. With Docker running on your Pi, one docker run gets you there:
docker run -d --restart=always -p 3001:3001 -v uptime-kuma:/app/data --name uptime-kuma louislam/uptime-kuma:1Or if you prefer Docker Compose (which you should, for easier management), a two-service compose file keeps everything tidy. Hit http://your-pi-ip:3001 in a browser, create your admin account, and you’re in. The dashboard loads immediately with a “Add New Monitor” button front and center.
For public access to your status page, I’d recommend pairing it with Cloudflare Tunnels — free, no port forwarding needed, and your status page is reachable from anywhere without exposing your home IP.
Tip 4 — Configure Telegram Notifications First
Uptime Kuma supports 90+ notification channels. I’ve tried several and Telegram is the best: instant delivery, reliable, works on every device, and you can create a private bot in 5 minutes via BotFather. When a service goes down, I get a message on my phone within 60 seconds. When it comes back up, I get another one. That two-way loop means I always know the current state of my stack without refreshing a dashboard.
To set it up in Uptime Kuma: Settings → Notifications → Add Notification → Telegram. Paste your Bot API Token and Chat ID. Test it. Done. You can also configure Slack, Discord, PagerDuty, and email — or stack multiple channels for critical services so alerts hit you more than one way.
Tip 5 — Monitor Beyond Just HTTP
Most people set up HTTP monitors for their web services and stop there. But Uptime Kuma can do a lot more in a homelab context. Here’s what I monitor beyond basic URLs:
- TCP port checks on my Plex server (port 32400) and SSH (port 22)
- Docker container heartbeats — push monitors that containers ping on a schedule, so I know if a container has stopped running even if its port looks open
- SSL certificate expiry — alerts 30 days before any cert expires so I’m not scrambling at the last minute
- DNS record checks — confirms my domain is resolving to the right IP after any DNS changes
- Ping monitors on my router and NAS — if those go dark, everything else is probably down too
Build a layered monitoring stack and you’ll know exactly what failed and roughly when — not just that “something’s wrong.”
Quick Answers: Uptime Kuma FAQ
Doesn’t monitoring from the same server defeat the purpose? Partly — if the whole box dies, the monitor dies with it, which is why the pros run their status checker on separate hardware. The homelab-pragmatic version: run Uptime Kuma on a Pi or second box if you have one, and if you don’t, it still catches the far more common failures — a single container crashing, a service hanging, a cert expiring — which is 90% of what actually goes wrong.
How aggressive should check intervals be? Sixty seconds is plenty for a homelab; twenty seconds for the one or two services the family notices instantly. Faster intervals mostly generate load and notification anxiety, not insight. The art of a good uptime kuma homelab setup is tuning it until every alert you receive is one you act on.
What’s the one notification setup worth doing properly? Route alerts somewhere you’ll see them but can silence — a dedicated Telegram or Discord channel beats email (ignored) and SMS (annoying). And set up the retry threshold so one flaky check doesn’t page you: alert on the second consecutive failure, not the first blip.
The Takeaway
An uptime kuma homelab setup takes about 15 minutes to get running and pays dividends every time it catches a service outage before you would have noticed manually. It’s one of those tools where once you have it, you can’t imagine running a homelab without it. Free, lightweight, and shockingly full-featured for a side project.
Pair it with a dedicated Raspberry Pi and a UPS, and you’ve got a monitoring stack that actually survives the scenarios where you need monitoring most. Start with Telegram notifications, add monitors for everything you care about, and build out a public status page your household can actually use.
What are you monitoring in your homelab? Drop your setup in the comments — I’m always looking for monitor types I haven’t thought to add yet. An uptime kuma homelab setup is thirty minutes of work for years of knowing before anyone else does.