XDA ran a piece this week about the specific way Docker containers can fail without crashing. A health check passes because the process is still up, but the app inside is wedged. Users see it before you do. The fix on the container side is a real health-check probe, not a process check. The fix on the operator side is a phone that tells you when something is off before a friend or a partner texts you.
The good news is that Android now has a proper set of home-lab observability apps. Not just companion viewers for a specific dashboard. Real alerting, real dashboards, real controls. What follows is seven Android apps we run against a modest home lab (Synology NAS, Proxmox mini-PC, three Raspberry Pi 5s, a small Kubernetes cluster) with actual alert-worthy incidents in the last month. Ranked by how often each one caught the problem first.
What matters when monitoring a home lab from a phone
- Live health beyond process state. The tool should check what the service does, not just whether its PID is still around.
- Push notifications that survive Doze mode. Android’s aggressive battery management kills poorly-configured alerting apps. First-party FCM push is more reliable than long-polling.
- Bidirectional controls. Restart a container, reboot a Pi, disable a broken sensor without opening a laptop.
- Federation. One phone screen for many services beats one app per stack.
- Auth that survives a session reboot. Long-lived tokens, biometric re-auth, and no “log in again” prompts every morning.
- Battery cost. A background-daemon app that polls every 30 seconds will eat 8% of a phone’s day. Event-driven and push-based apps stay near zero.
Quick comparison
| App | Best for | Free plan | Starting price | Standout feature |
|---|---|---|---|---|
| Home Assistant | Physical devices + services | Yes, open source | Free, HA Cloud about 7 USD/month | 3,000-plus integrations |
| Uptime Kuma | Endpoint uptime dashboard | Yes, open source | Free | Pretty status page bundled |
| Portainer Companion | Docker and Kubernetes fleets | Yes | Free with Portainer CE | Container start/stop from the phone |
| Grafana | Deep dashboards with history | Yes | Free with Grafana OSS | Time-series drill-down |
| Netdata | Real-time metrics per host | Yes | Free tier, paid Cloud tier | Per-second resolution out of the box |
| Prometheus Alertmanager | Rule-based alerting | Yes | Free | Sophisticated alert routing |
| Docker Manager | Lightweight container control | Yes | About 3 USD Pro | Runs against Docker socket directly |
The apps
1. Home Assistant — Best all-in-one home lab watchtower
Home Assistant started as a smart-home hub and quietly became the best free home-lab dashboard on Android. The companion app talks to a self-hosted Home Assistant instance over LAN or a reverse proxy. Push notifications, sensor charts, service restarts, and long-running automations all live in one place. Docker container states, network switch power draw, and Uptime Kuma alerts can all be surfaced inside a single Home Assistant view.
Where it falls short: The setup effort is real. Home Assistant is best when you commit to it as the hub. Half-committing means the app duplicates work your other tools already do.
Pricing:
- Free: everything, forever
- Paid: Home Assistant Cloud at about 7 USD/month for remote access without a reverse proxy
Platforms: Android, iOS, web
Download: Home Assistant on Aptoide · Google Play
Bottom line: The best pick if you want one app and one dashboard for the whole house and the whole lab.
2. Uptime Kuma — Best endpoint uptime dashboard
Uptime Kuma runs as a container on your NAS and pings your services. The web UI is polished, and the app-side experience on Android is a mobile-friendly view of the same dashboard, plus push notifications when a check fails. Watching HTTP endpoints, ports, DNS records, and container health checks in one place makes it the pick for the “did anything break” question at 8am.
Where it falls short: Not a general-purpose observability tool. It knows if a service is up, not why. No first-party Android app; you use the web UI or a third-party client.
Pricing:
- Free: MIT open source
- Paid: none
Platforms: Web-first, mobile-friendly
Download: Uptime Kuma mobile client on Aptoide — community client for the Kuma API
Bottom line: Pair it with a real alerting channel (ntfy, Pushover, or Home Assistant) and it becomes the single “up or down” pane.
3. Portainer Companion — Best Docker and Kubernetes control
Portainer Companion is a third-party Android client for the Portainer API. That matters because Portainer CE runs in every home lab that has Docker or Kubernetes, and the app lets you restart a container, exec into a shell, or wipe a stuck volume from the phone. On-call rotations that boil down to “restart the container” are covered by this alone.
Where it falls short: Third-party status means the API surface can drift. Some newer Portainer features arrive later on the Android side. The UI is functional rather than pretty.
Pricing:
- Free: full functionality against Portainer CE
- Paid: none for the app
Platforms: Android, iOS
Download: Portainer Companion on Aptoide · Google Play
Bottom line: Install this if Portainer is already in the stack.
4. Grafana — Best deep dashboards
Grafana on mobile is still a viewer more than a builder, but its viewer is very good. Once you have set up dashboards on the desktop, the Android app makes them scannable. Alerting hooks into any provider that speaks a webhook. Loki and Prometheus integrations mean logs and metrics land in the same tap.
Where it falls short: Building dashboards on a phone is unpleasant. Alerting configuration must happen on the web side. The app is best on a tablet, not a phone screen.
Pricing:
- Free: Grafana OSS, self-hosted
- Paid: Grafana Cloud from about 30 USD/month for larger setups
Platforms: Android, iOS, web
Download: Grafana on Aptoide · Google Play
Bottom line: Pick this if the lab already has metrics history worth zooming into.
5. Netdata — Best real-time metrics per host
Netdata ships an agent per host that pushes per-second metrics into a dashboard. The Android app displays those dashboards, shows per-service anomalies, and pushes alerts when a host misbehaves. The free tier is generous. The Cloud tier adds cross-host correlation and central alerting.
Where it falls short: Cross-host correlation without the Cloud tier requires a self-hosted parent Netdata that many home labs will skip. Alert deduplication is thinner than Alertmanager.
Pricing:
- Free: full agent and single-host dashboards
- Paid: Netdata Cloud tiers from about 4 USD/month per active user for cross-host and long retention
Platforms: Android, iOS, web
Download: Netdata on Aptoide · Google Play
Bottom line: Pick this if you want per-second visibility and are fine per-host rather than fleet-wide.
6. Prometheus Alertmanager — Best rule-based alerting
Prometheus and Alertmanager together are the alerting spine of most serious home labs. The Android side is a viewer plus a routing target: Alertmanager pushes to Pushover, Discord, ntfy, or a webhook, and the phone gets the ping. Rule complexity is the point. “Only alert if the container has been unhealthy for five minutes AND the CPU is over 80% for the same period” is trivial here and impossible in most competitors.
Where it falls short: Alertmanager is not “install and go”. Prometheus scrape configs are their own topic. If you have never written a PromQL query, plan a weekend.
Pricing:
- Free: Apache-licensed open source
- Paid: none
Platforms: Web-first; Android via Grafana or dedicated Alertmanager clients
Download: No first-party Android app; alerts arrive via configured routing.
Bottom line: Pick this when the false-positive rate on the simpler alerting tools has become unacceptable.
7. Docker Manager — Best lightweight container control
Docker Manager is a simpler alternative to Portainer Companion. It talks straight to the Docker socket (over SSH or a TCP endpoint you expose to LAN) and lets you list, start, stop, and inspect containers. No Portainer server required. For a small home lab, the lower operational overhead is the pitch.
Where it falls short: No Kubernetes support. Exposing the Docker socket over the network is a security choice that needs a hardening layer (Tailscale, VPN, TLS + client certs).
Pricing:
- Free: full features
- Paid: about 3 USD one-time Pro to remove ads and unlock multi-endpoint
Platforms: Android
Download: Docker Manager on Aptoide · Google Play
Bottom line: Pick this when the lab is Docker-only and you would rather not run Portainer at all.
How to choose
- If you want one dashboard and one app for house and lab: Home Assistant.
- If the daily question is “is anything down”: Uptime Kuma.
- If restarting a container is the on-call: Portainer Companion.
- If deep dashboards with history matter: Grafana.
- If per-host real-time metrics are the point: Netdata.
- If alert rules need to be smarter than a threshold: Prometheus Alertmanager.
- If Docker is the whole lab and simple is better: Docker Manager.
Stay on whatever web dashboards you already have if a phone view is not going to make you act faster. The value of these apps is not looking at graphs. It is fixing a thing in the ten minutes between noticing it and forgetting it.
FAQ
Do these apps drain the phone battery? Well-configured push-based apps (Home Assistant, Netdata Cloud, Uptime Kuma via ntfy) stay near zero. Polling-heavy apps (some third-party clients) can eat 8-10% a day if configured to check every 30 seconds.
Which one alerts about a Docker health check failure? Uptime Kuma if you point it at a health endpoint. Prometheus with cAdvisor scraping the container. Home Assistant with a Docker integration also works.
Can I use these without a VPN? Yes if the service is reachable via a reverse proxy with TLS. Tailscale, WireGuard, or Cloudflare Tunnel are common patterns for making these apps work from anywhere.
What about NAS-specific apps like Synology DSM or Truenas? Those exist and are useful for NAS-native monitoring, but they only cover one box. The apps above see the whole lab.
Is Home Assistant overkill for just monitoring? Not really. The dashboard alone is worth it, and it grows with the lab. But if you are strictly monitoring services and not smart-home devices, Uptime Kuma is a lighter starting point.
