Product / Monitoring
Server Monitoring Dashboard That Shows You What Broke
See live CPU, memory, disk, and network for every VPS in one place. Scrub through 24 hour, 7 day, and 30 day graphs. Get paged on Slack, Discord, email, Telegram, or SMS the moment a threshold slips, with the last ten minutes of context already attached.
No agent to install. Works with any Ubuntu 22.04 or 24.04 server over SSH.
What you get in one dashboard
Live metrics every 30 seconds. Historical graphs with 1m, 5m, 1h, and 1d rollups. Uptime checks from three regions. Alerts to six channels with smart debouncing. Anomaly detection that learns your baseline. Process and systemd service health. Prometheus export if you want to plug it into your own stack.
Real time metrics
CPU, memory, disk, and network, every 30 seconds
The four signals that actually matter, polled often enough to catch a spike and displayed cleanly enough to read on a phone. No setup, no agent, no extra port to punch through the firewall.
CPU load
Per core usage, one minute and five minute load averages, top processes by CPU. Catches a runaway worker before your users do.
Memory + swap
Used, cached, and swapped memory in one chart. Warns you before the OOM killer wipes out PHP FPM at 3 AM.
Disk + inodes
Per volume usage and read and write throughput. Tracks inodes too, so a runaway log directory cannot sneak up on you.
Network I O
Inbound and outbound bandwidth per interface, with packet rate and error counters. Spot a DDoS or a misconfigured cron in seconds.
Historical graphs
Scrub through 24 hour, 7 day, and 30 day views
Every metric stores rollups at one minute, five minute, one hour, and one day resolutions. Compare two servers side by side. Compare the same server across two weeks. Spot slow drift long before it turns into an incident.
- Smooth 1m resolution for the first seven days
- 5m rollups for the next 30 days on Pro
- 1h rollups kept for 60 more days on Pro
- One click compare with any other server you own
- Export any time range as CSV or PNG
Retention at a glance
- Free tier
- 7 days at 1m resolution
- Pro tier
- 90 days, rolled up over time
- Enterprise tier
- Up to 3 years, S3 export
- Resolutions
- 1m, 5m, 1h, 1d
Uptime checks
Probes from three regions, every 30 seconds
One failed region sends a warning with a traceroute. Two failed regions flip your public status page to degraded automatically. Keep customers informed without writing a single line of status page code.
Three global regions
Frankfurt, Virginia, and Singapore probe your endpoints every 30 seconds. We publish the IP ranges so you can allow list us.
HTTP, TCP, and port checks
Check an HTTP URL, a raw TCP port, or a TLS handshake on any port. Use HEAD by default, switch to GET when your upstream needs it.
Public status page
A hosted status page at status.yourdomain.com with live metrics, incident history, and email subscriber sign up, wired up in one click.
Alerts that respect your evening
Slack, Discord, email, Telegram, SMS, or webhook
A tool that pages you every five minutes trains you to ignore it. Molixa Forge defaults to fewer, smarter alerts, with escalation you tune per server and per channel.
- Debounced by default, five straight polls over threshold before firing
- Deploy aware, monitoring alerts pause while a deploy is running
- Escalation after N minutes, call the next person on the rotation
- Quiet hours per team member, with P0 override for real emergencies
- Every alert includes a sparkline and a link straight into the incident view
Slack + Discord
Rich message with metric sparkline, severity color, and a one click link to acknowledge.
SMS + Telegram
SMS via Twilio or Telegram bot, tuned for criticals so you get paged only when you need to be.
Webhook
Plain JSON POST to any endpoint. Wire it into PagerDuty, Opsgenie, or your own on call tool.
Always on fallback with full incident details and a link to the dashboard. Never misses a fire.
Anomaly detection
Catch weird, even when no threshold fires
Static thresholds miss the slow leaks. The server monitoring dashboard learns the baseline for each server and pings you when a metric drifts too far from its own normal, without extra tuning.
Baseline learning
The system builds a 14 day rolling profile per metric per server. Weekday patterns, night lulls, and monthly batch jobs all register as normal, so you stop getting paged for your own Sunday cron.
Drift detection
A slow memory leak that adds 80 MB a day never crosses a static threshold but still crashes the box on day 12. Drift detection catches that on day 4 and flags it, with a sparkline and the suspected process.
Smart grouping
When five servers spike at once, you get one grouped alert, not five. When one server spikes twice in ten minutes, you get one grouped alert with both events, so the channel stays readable.
Process and service health
Watch the services your app actually depends on
Nginx, PHP FPM, MySQL, PostgreSQL, Redis, PM2, Docker containers, any systemd unit you point at. The dashboard shows current state, memory per process, CPU per process, and flips a red dot the moment a unit goes inactive.
- Top 10 processes by CPU and memory, refreshed every minute
- Systemd service state for the units you pick
- Restart count, so a flapping service does not hide
- Log tail in the browser for every watched service
- Per site health rollup, so you see which app is affected
Built for the stacks you run
- Nginx
- Apache
- PHP FPM
- Node.js + PM2
- MySQL + MariaDB
- PostgreSQL
- Redis
- Docker
- Laravel queues
- WordPress cron
Want to pair monitoring with an assistant that actually acts on the alerts? See the Molixa AI assistant or read our take on the best free server management panels for 2026.
FAQ
Monitoring questions, answered
How often does Molixa Forge collect server metrics?
The server monitoring dashboard polls CPU, memory, disk, and network every 30 seconds over your existing SSH connection. There is no agent to install and no extra port to open. For process and service checks, the interval drops to 15 seconds on Pro, so you see spikes and restarts as they happen.
How long are historical metrics kept?
Free keeps seven days at one minute resolution. Pro keeps 90 days with one minute for the first week, five minute rollups for the next month, and one hour points after that. Enterprise keeps up to three years and can export cold data to your own S3 bucket on a one click toggle.
Which alert channels are supported out of the box?
Slack, Discord, Telegram, email, SMS via Twilio, and plain webhooks to any endpoint you control. Each channel has its own schedule and severity rules, so warnings can hit Slack while criticals page your on call rotation by SMS. Quiet hours are honored per channel and per team member.
How fast are uptime checks, and where do probes run from?
Uptime checks run every 30 seconds from three regions by default, Frankfurt, Virginia, and Singapore. One failed region pings the relevant on call with a traceroute. Two failed regions flip your public status page to degraded automatically, so users see the same truth you do.
Can I export metrics to my own Prometheus or Grafana stack?
Yes. The server monitoring dashboard exposes a Prometheus format endpoint per server, so you can scrape from your own Prometheus and chart in your own Grafana. Molixa Forge stays useful for alert routing and log tailing while the raw data lives wherever you want it.
How do I stop noisy alerts during deploys or maintenance?
Every alert is debounced by default, so a metric must stay over threshold for five straight polls before it fires. Deploys automatically pause related alerts until the deploy completes. You can also open a maintenance window with a click and suppress everything for a set duration.
See your servers clearly
Free tier includes live metrics, seven day history, and email alerts for one server. No credit card.