Watch your boxes.
Sleep at night.
Server monitoring that does not cost a fortune. Get CPU, memory, disk, and network metrics with one command. Disk predictions, uptime SLA tracking, maintenance windows, cron job heartbeat monitoring, process monitoring, uptime monitoring, and alerts via Slack, Discord, or webhooks.
curl -sL boxwatch.app/install.sh | bash -s YOUR_KEY
How It Works
From zero to monitored in under a minute
Copy the Install Command
Grab your unique install command from the dashboard. It includes your API key.
Paste on Your Server
Run the command via SSH. The agent installs itself and starts reporting instantly.
Relax & Get Alerts
Metrics flow in automatically. Set up alerts and sleep knowing you will be notified.
The agent script is open source. Review exactly what it collects before installing.
Datadog: $375/movsBoxWatch: $13/mo
Same metrics. 30x less cost. For 7 servers.
Uptime Monitoring
URL and service monitoring from your own infrastructure.
Probe HTTP endpoints, TCP ports, and TLS cert expiry from the BoxWatch agents you've already installed. Multi-region monitoring as a side effect of your existing deployment. Internal-network TCP checks (verify Redis is reachable from web-01) — something UptimeRobot and Better Stack literally cannot do. Linux-only via agent v2.1.
Process Monitoring
Watch the processes that matter, not just the host.
Track CPU, memory, and uptime for nginx, postgres, your queue workers — anything that runs as a named process. Get notified when a process dies, restarts, exceeds a CPU threshold, or eats too much memory. Aggregate by name across worker pools. Linux-only via agent v2.0. Included on every plan — 10 watched processes per server on Hobby, 50 on Pro, 100 on Team, unlimited on Scale.
Cron Heartbeat Monitoring
Monitor your scheduled jobs without an agent.
Get notified the moment a scheduled job misses its window, fails, hangs, or returns a non-zero exit code. Drop-in ping URLs for /start, /success, and /fail/:code work with any cron, systemd timer, Kubernetes CronJob, GitHub Actions workflow, or one-off script. Transition-based alerts mean no storms — you hear about state changes, not every cycle. Included on every plan — 20 checks on Hobby, 100 on Pro, unlimited on Team and Scale.
60-Second Setup
One curl command. No YAML configs. No agent ecosystem to learn. Just paste and go.
Custom API Endpoints
Create filtered views of your data. Each endpoint gets a unique URL. Pull into your own tools.
TV Dashboard Mode
Full-screen dashboards for your NOC or office wall. Dark mode. Auto-refresh. Just add a screen.
Slack, Discord & Email
Get instant notifications where your team already is. Server down? CPU spiking? Know immediately via webhook or email.
Uptime Badges
Shields.io-style badges for your README. Show off your 99.9% uptime. Auto-updates every 5 minutes.
Public Status Pages
Branded status pages for your customers. Show real-time uptime, incidents, and 90-day history. No extra tools needed.
Server Groups
Organize servers into named groups. Filter, compare, and manage by group. Perfect for agencies managing client infrastructure.
Disk Full Prediction
AI-powered disk usage forecasting. Get alerted days before your disk fills up, not after. Linear regression on 7 days of data.
Uptime SLA Tracking
Track 24h, 7d, 30d, and 90d uptime percentages. Maintenance windows excluded from calculations. Color-coded at a glance.
Server Comparison
Compare CPU, memory, or disk across all servers on one chart. Spot the outlier instantly.
Maintenance Windows
Schedule downtime that automatically suppresses alerts and excludes from uptime calculations. No more false alarms during updates.
Webhook Alerts
Send alerts to any URL. Custom headers for auth. Works with PagerDuty, Zapier, n8n, or your own scripts. Test button included.
Weekly Digests
Short weekly email summary: server health, uptime, alerts, disk warnings. Monthly detailed report also available. Toggle in settings.
Privacy-First Monitoring
BoxWatch only collects system performance metrics (CPU, memory, disk, network). We never access your files, logs, databases, or sensitive data. The agent script is open source so you can verify exactly what runs on your servers.
How We Compare
BoxWatch vs the alternatives
| Feature | BoxWatch $13/mo | Datadog $15+/host | UptimeRobot $7/mo | Self-Hosted Free |
|---|---|---|---|---|
| CPU, Memory, Disk Metrics | ✓ | ✓ | ✗ | ✓ |
| One-Command Install | ✓ | ~ | ✗ | ✗ |
| TV Dashboard Mode | ✓ | ✓ | ✗ | ~ |
| Slack/Discord Alerts | ✓ | ✓ | ✓ | ~ |
| Public Status Pages | ✓ | $ | ✓ | ✗ |
| No Server to Manage | ✓ | ✓ | ✓ | ✗ |
| Disk Full Prediction | ✓ | ✓ | ✗ | ~ |
| Maintenance Windows | ✓ | ✓ | ✗ | ~ |
| Webhook Alerts | ✓ | ✓ | ~ | ✓ |
| Uptime SLA Tracking | ✓ | ✓ | ✓ | ~ |
| Cron Job Heartbeats | ✓ | ✓ | ~ | ~ |
| Process Monitoring | ✓ | ✓ | ✗ | ~ |
| Multi-region Probes (via your agents) | ✓ | ~ | ✗ | ~ |
| Price for 7 Servers | $13/mo | $105+/mo | No metrics | Your time |
✓ = Included ~ = Partial/Complex ✗ = Not Available $ = Extra Cost
Simple, Predictable Pricing
No per-metric charges. No surprise bills. No sales calls.
- ✓ 2 servers
- ✓ 1 alert rule
- ✓ 20 cron checks
- ✓ 20 uptime checks
- ✓ 10 watched processes per server
- ✓ 60-min intervals
- ✓ 24h retention
- ✓ 1 API endpoint
- — No TV dashboards
- ✓ 7 servers
- ✓ 5 alert rules
- ✓ 100 cron checks
- ✓ 100 uptime checks
- ✓ 50 watched processes per server
- ✓ 5-min intervals
- ✓ 7-day retention
- ✓ 3 TV dashboards
- ✓ 10 API endpoints
- ✓ Weekly digests
- ✓ Maintenance windows
- ✓ 25 servers
- ✓ 20 alert rules
- ✓ Unlimited cron checks
- ✓ Unlimited uptime checks
- ✓ 100 watched processes per server
- ✓ 5-min intervals
- ✓ 30-day retention
- ✓ 6 TV dashboards
- ✓ 25 API endpoints
- ✓ Weekly digests
- ✓ Maintenance windows
- ✓ 100 servers
- ✓ 50 alert rules
- ✓ Unlimited cron checks
- ✓ Unlimited uptime checks
- ✓ Unlimited watched processes
- ✓ 1-min intervals
- ✓ 90-day retention
- ✓ 25 TV dashboards
- ✓ Unlimited endpoints
- ✓ Weekly digests
- ✓ Maintenance windows
Quick Questions
Need more? Check out our full FAQ
How does the agent work?
The BoxWatch agent is a lightweight bash script that collects system metrics (CPU, memory, disk, etc.) and sends them to our API every few minutes. It runs via cron and uses minimal resources. No root access to your files, logs, or data.
Can I monitor my cron jobs?
Yes. Create a cron check in BoxWatch, copy the ping URL, and have your job hit it on success (and optionally /start and /fail/:exit_code too). BoxWatch alerts you if a job misses a run, fails, hangs, or runs too long. Generous limits even on the free tier — 20 cron checks on Hobby, 100 on Pro, unlimited on Team and Scale.
Can I monitor specific processes?
Yes, on Linux hosts. On each server's detail page, click "+ Watch new" and enter a process name (the exact command name pgrep -l shows — e.g., nginx, postgres, redis-server). BoxWatch tracks aggregate CPU/memory across all matching instances and alerts you when the process dies, restarts, exceeds a CPU threshold, or exceeds a memory threshold. Generous limits on every plan — 10 per server on Hobby, 50 on Pro, 100 on Team, unlimited on Scale.
Can I monitor URLs and external services?
Yes. Create an uptime check (HTTP, TCP, or TLS cert expiry), pick which of your BoxWatch agents should probe it, and get alerted when the check fails. Because the probes run from YOUR agents, you get multi-region monitoring naturally — and you can probe internal network targets (Redis on private IPs, internal services) that centralized monitoring services can't reach.
What if I exceed my server limit?
Additional servers beyond your plan limit will be marked as inactive and won't receive metrics. You can upgrade your plan anytime to activate them instantly. No surprise charges.
How do I remove a server?
Run our uninstall command on your server: curl -sL boxwatch.app/uninstall.sh | bash then delete it from your dashboard. Simple.
Is there a free tier?
Yes! The Hobby plan is free forever with 2 servers, 1 alert rule, and 24-hour data retention. No credit card required to start. Upgrade only when you need more.
Ready to watch your boxes?
Start monitoring 2 servers for free. No credit card required. Upgrade when you need more.