Your servers, your terminal,
nothing exposed
Nexoder is a rack-style dashboard for your Linux VPS fleet that lives entirely in your terminal. It speaks plain SSH with the keys you already have.
| Service | Status | Uptime | CPU% | Mem |
|---|---|---|---|---|
| caddy.service | active | 84d 6h 11m | 0.2 | 38.1 MB |
| docker.service | active | 84d 6h 11m | 1.8 | 412.7 MB |
| postgresql.service | active | 12d 4h 2m | 0.6 | 238.1 MB |
| fail2ban.service | active | 84d 6h 11m | 0.1 | 34.9 MB |
| certbot.service | failed | — | — | — |
| ★ | PID | User | CPU% | MEM% | RSS (MB) | Command |
|---|---|---|---|---|---|---|
| ★ | 2114 | root | 1.8 | 5.2 | 427 | dockerd |
| 4821 | postgres | 0.6 | 9.1 | 745 | postgres: checkpointer | |
| 9932 | vdvoem | 2.4 | 4.8 | 391 | gunicorn: worker (+3 more) | |
| 1183 | root | 0.3 | 2.9 | 238 | python3 | |
| 612 | caddy | 0.2 | 0.5 | 38 | caddy |
Three zeros your monitoring stack can't offer
Every mainstream monitoring tool wants to live on your server, open a port, or ship your metrics to someone's cloud. Nexoder refuses all three.
agents installed
Nothing to deploy, update, or clean up on the server. Nexoder runs on your machine and reads metrics through one batched SSH command per refresh.
ports exposed
No web dashboard listening on :19999, no /metrics endpoint to firewall. Your attack surface stays exactly what it was: sshd, which you already run.
bytes leave your machine
No cloud, no accounts, no telemetry, no phone-home. Server records live in an encrypted SQLite vault in your local config directory.
The whole app, four keystrokes deep
Unlock, pick a box in the rack, read the dashboard, audit what auto-starts at boot. Full keyboard navigation, no mouse required.
| Status | Label | Address | Checked |
|---|---|---|---|
| ● online | selesta | 84.23.101.7:22 | 14:32:07 |
| ● online | fata-morgana | 45.132.18.94:22 | 14:32:07 |
| ● online | edge-fra-1 | 91.98.12.44:22 | 14:32:07 |
| ● checking… | build-runner | 10.0.4.17:22 | 14:31:52 |
| ● offline | staging-old | 92.51.44.130:22 | 14:32:07 |
| + | + Add New Server |
selesta Status: ● online IP: 84.23.101.7:22 Host: selesta OS: Ubuntu 24.04.2 LTS Uptime: 84d 6h 12m CPU: 12.4% (4 cores) RAM: 63% 5.0 GB / 8.0 GB Disk: 81% 129.4 GB / 160.0 GB Services: 4/5 Processes: 100 snapshot: 12s ago
| Unit | Active | User | Preset |
|---|---|---|---|
| caddy.service | active | caddy | enabled |
| docker.service | active | root | enabled |
| fail2ban.service | active | root | enabled |
| postgresql.service | active | postgres | enabled |
| ssh.service | active | root | enabled |
| unattended-upgrades.service | active | root | enabled |
A full ops cockpit in one TTY
Built for people who run 1 to 30 boxes and refuse to stand up Grafana for their own infra.
Multi-server rack view
Every box in one table with a live reachability LED and a last-known snapshot. Stop cycling through ssh sessions to answer "is everything up?"
Batched refresh
The whole dashboard - OS info, load, CPU, RAM, disk, services, processes - arrives in a single bash round-trip every 15 seconds.
Process table + pinning
Top 100 processes sortable by CPU or memory. Pin the ones you care about; they float to the top and persist across reconnects.
One-key service restart
Hit Enter on a process and Nexoder resolves its systemd unit via /proc cgroups and restarts it. The only write action, and an explicit one.
Enabled-units audit
See everything that auto-starts at boot: unit, active state, running user, vendor preset. Find the cron-miner before it finds you.
Encrypted local vault
One master password unlocks all your server records. Argon2id key derivation, AES-GCM at rest, key held in memory only.
Paranoid by construction
Nexoder is a desktop tool. There is no server component to breach, no SaaS custodian of your data, no third party in the loop.
Monitor the fleet, keep the perimeter
Public release is in the works.