Fleet Command Center: Master Architecture & Operational Guide

Target System: omni-yo-221 (Oracle Cloud Infrastructure Always Free AMD Instance)
Created: October 2026
Primary Purpose: Single-point-of-contact lightweight command center to manage, update, inspect, and SSH into a fleet of remote servers with zero idle memory overhead.


⚠️ CRITICAL NOTICE FOR FUTURE AI ASSISTANTS

READ THIS FIRST BEFORE PROPOSING OR INSTALLING SOFTWARE ON THIS SERVER:

  1. Hardware Ceiling: This is an Oracle Cloud VM.Standard.E2.1.Micro instance:
  2. Physical RAM: ~951 MiB total (with 1.0 GiB swap).
  3. Compute: 1 OCPU / 2 vCPUs (fractioned AMD EPYC 7551).
  4. Storage: 45 GB SSD.
  5. Anti-Bloat Mandate:
  6. DO NOT propose or install heavy containerized management suites (e.g., Cockpit, Portainer, heavy Node.js dashboards, Uptime Kuma, Linkwarden, or multiple Docker runtimes).
  7. This server routinely operates near ~500 MB memory. Stacking heavy runtimes triggers the Linux Out-Of-Memory (OOM) killer and freezes the server via disk swap thrashing.
  8. The Fleet System: Multi-server management is handled by a native, dependency-free Python CLI tool (/usr/local/bin/fleet, aliased to c). Do not replace it with an external platform unless explicitly instructed by the user.
  9. SSH & Keys: All remote server connections use the dedicated ed25519 keypair at ~/.ssh/id_ed25519. Do not overwrite or delete this key.

1. System Architecture Overview

                                +-----------------------------+
                                |  Oracle Cloud VPS (1GB RAM) |
                                |       Host: omni-yo-221     |
                                +--------------+--------------+
                                               |
                                     [/usr/local/bin/fleet]
                                     (Python 3 Standard Lib)
                                               |
                      +------------------------+------------------------+
                      |                                                 |
             [Interactive Mode]                                [Batch Execution]
         fzf + OpenSSH (Multiplexed)                     ThreadPoolExecutor (Parallel)
                      |                                                 |
       +--------------+--------------+                   +--------------+--------------+
       |                             |                   |                             |
       v                             v                   v                             v
[ Remote Server 1 ]          [ Remote Server 2 ]   [ Remote Server 1 ]          [ Remote Server 2 ]
(Interactive Shell)          (Interactive Shell)   (apt update / cmd)           (apt update / cmd)

Component Details

  • CLI Executable: /usr/local/bin/fleet (and symlinked /usr/local/bin/c)
  • Source Code: /home/ubuntu/fleet.py
  • Inventory File: ~/.fleet/inventory.json
  • SSH Key: ~/.ssh/id_ed25519 (public key: ~/.ssh/id_ed25519.pub)
  • SSH Multiplexing Sockets: ~/.ssh/sockets/ (configured in ~/.ssh/config)
  • Idle RAM Consumption: 0.00 MB (process only runs when a command is invoked).

2. Command Reference for the Human Operator

2.1 Quick Shell Aliases

Alias Command Purpose
c fleet ssh Open interactive fuzzy menu to pick a server and log in
c <name> fleet ssh <name> Instantly log into <name>
fs fleet status Live parallel health table (RAM, Disk, Load, Uptime, Containers)
fup fleet update Run sudo apt update && upgrade -y && autoremove -y across all servers in parallel
fup <name> fleet update <name> Run full update and autoremove on a single server
frun "<cmd>" fleet run "<cmd>" Run any bash command across all servers in parallel
fleet logs <srv> fleet logs <srv> [unit] Stream live logs for a container or systemd unit
Web Dashboard https://stateoftheshit.com Live dark-mode telemetry (RAM, Disk, Uptime, Containers)

2.2 Adding and Managing Servers

Step 1: Copy Command Center Public Key to Remote Host

View this machine's public key anytime:

fleet pubkey

Key value:

ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIBGxf3wP7+lIpn0b7OLtuCjS98DW6dk97SjTVHsLeqKP fleet-command-center

Add this key into ~/.ssh/authorized_keys on your other servers.

Step 2: Add to Inventory

# Direct syntax:
fleet add <server_alias> <ip_or_host> [user] [port] [group]

# Examples:
fleet add prod-web 192.168.1.50 root 22 prod
fleet add backup-box 51.15.20.10 ubuntu 22 backups

# Or run interactively (prompts you step-by-step):
fleet add

Step 3: View or Remove Servers

fleet list           # View all configured servers
fleet remove <name>  # Remove a server from inventory

2.3 Interactive Server Jump

Type c or fleet: 1. An interactive fzf search menu appears showing your servers, usernames, IPs, and tags. 2. Fuzzy-type any part of the name (e.g. web, prod, 192). 3. Press Enter → You are immediately logged into that server.


2.4 Running Commands Across Your Fleet

# Check disk and RAM across all servers:
fleet run "df -h /; free -m"

# Check Docker status everywhere:
fleet run "systemctl is-active docker"

# Target a specific group/tag:
fleet run -g prod "tail -n 20 /var/log/nginx/error.log"

# Target a single server:
fleet run -s prod-web "uptime"

2.5 Fleet-Wide Package Updates

# Update every registered server at once:
fleet update

# Update a single server:
fleet update prod-web

Automatically detects Debian/Ubuntu (apt), RHEL/Fedora (dnf), and Arch (pacman).


2.6 Live Health Check Probe

fleet status

Produces a real-time, parallel summary:

NAME             STATUS     UPTIME                 LOAD             MEMORY         ROOT DISK           
--------------------------------------------------------------------------------------------------
prod-web         ONLINE     up 21 days, 4 hours    0.14 0.09 0.05   410M/2048M     24G free (32% used) 
backup-box       ONLINE     up 8 days, 12 hours    0.02 0.01 0.00   180M/1024M     120G free (40% used)
staging-1        OFFLINE    -                      -                -              Connection timed out

3. Technical Implementation & Internals (For Future AI Chats)

If the user asks a future AI chat to extend, modify, or troubleshoot fleet:

3.1 Inventory Storage Schema

Location: /home/ubuntu/.fleet/inventory.json

{
  "server_name": {
    "host": "1.2.3.4",
    "user": "root",
    "port": 22,
    "group": "prod",
    "auth_type": "key"
  }
}

3.2 SSH Performance Tuning

Location: /home/ubuntu/.ssh/config

Host *
    StrictHostKeyChecking accept-new
    ControlMaster auto
    ControlPath ~/.ssh/sockets/%r@%h-%p
    ControlPersist 10m
    ServerAliveInterval 30
    ServerAliveCountMax 3
    IdentityFile ~/.ssh/id_ed25519
  • Why ControlMaster? Sockets are kept in ~/.ssh/sockets/ for 10 minutes. When running batch commands (fleet status or fleet run), parallel connections reuse open TCP sockets without renegotiating SSH handshakes every time, speeding up execution by 500-1000%.

3.3 Extending the fleet CLI

  • To modify or add functionality, edit /home/ubuntu/fleet.py.
  • Copy the updated file to /usr/local/bin/fleet: bash sudo cp /home/ubuntu/fleet.py /usr/local/bin/fleet sudo chmod +x /usr/local/bin/fleet
  • Rule: Do NOT introduce heavyweight third-party libraries (keep it pure standard library or use system tools like fzf).

Fleet Command Center Quick Reference

A zero-RAM, terminal-native central control tool installed at /usr/local/bin/fleet.


1. Setup Your Other Servers

Your command center SSH public key:

ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIBGxf3wP7+lIpn0b7OLtuCjS98DW6dk97SjTVHsLeqKP fleet-command-center

(View anytime by running fleet pubkey)

Adding a Server

fleet add <name> <ip_or_domain> [user] [port] [group]

Example:

fleet add web1 192.168.1.100 root 22 prod
fleet add vps2 51.15.20.30 ubuntu 22 dev

(You can also just run fleet add without arguments for an interactive prompt).


2. Instant Server Login

  • Run c or fleet:
  • Pops up an interactive fuzzy finder (fzf).
  • Type a couple of letters, hit Enter, and you're immediately logged into the server over SSH!
  • Direct jump: bash fleet ssh web1 # or c web1

3. Run Commands Across Servers

Run any command across all servers in parallel:

fleet run "df -h /; free -m"
fleet run "systemctl status docker"
fleet run "sudo reboot"

Run on a single server:

fleet run -s web1 "uptime"

Run on a specific group/tag:

fleet run -g prod "tail -n 20 /var/log/nginx/error.log"

4. One-Command Fleet Updates & Autoremove

Update package lists, upgrade packages, and autoremove unneeded packages across your fleet in parallel: * All servers at once: bash fup # or fleet update * Specific server: bash fleet update vault (Executes: sudo apt update && sudo apt upgrade -y && sudo apt autoremove -y).


5. Live Fleet Health & Web Dashboard

In the Terminal:

Check connectivity, uptime, load average, memory, disk space, and active containers across all machines:

fs
# or
fleet status

In Your Browser (Zero-RAM Static Web UI):

Open https://stateoftheshit.com on your phone, laptop, or tablet. * Live server status & ping * Uptime & load average * RAM & root disk progress bars * Active Docker containers (e.g. Vaultwarden, Linkwarden, etc.) * Auto-refreshes every 30 seconds


6. Stream Live Logs

Stream live container or systemd logs with one command:

fleet logs vault vaultwarden
fleet logs vault nginx
fleet logs agentdover

7. Shortcuts Summary

Command Shortcut What it does
fleet / fleet ssh c Interactive fuzzy server selector & login
fleet ssh <name> c <name> Direct SSH into <name>
fleet status fs Live status table (RAM, Disk, Uptime, Load, Containers)
fleet update fup Run apt update && upgrade -y && autoremove -y everywhere
fleet run "<cmd>" frun "<cmd>" Run shell command across all servers
fleet logs <srv> - Stream live logs for a container or service
fleet dashboard - Re-build web dashboard (auto-runs every 1 min)
Web Dashboard - Live browser view at https://stateoftheshit.com
fleet list - Print server inventory
fleet add - Register a new server
fleet remove <name> - Remove a server
fleet pubkey - Display command center public key

Server Specifications & System Status Report

Report Generated: 2026-10-09 17:47 UTC
Hostname: omni-yo-221
Host Platform: Oracle Cloud Infrastructure (OCI) Compute Instance
Public IP: 150.136.173.182
Private IP: 10.0.147.155/16 (ens3)


1. Hardware & System Specifications

1.1 Compute & Processor

  • CPU Model: AMD EPYC 7551 32-Core Processor
  • vCPUs: 2 (1 Socket, 1 Core per socket, 2 Threads per core)
  • Architecture: x86_64 (64-bit)
  • Hypervisor: KVM (QEMU Standard PC i440FX + PIIX)

1.2 Memory & Swap

  • Total Physical RAM: ~951 MiB (~1.0 GB)
  • Used: ~553 MiB
  • Buff/Cache: ~502 MiB
  • Available: ~397 MiB
  • Swap: 1.0 GiB (126 MiB used, ~897 MiB free)

1.3 Storage & Partitions

  • Primary Disk (/dev/sda): 46.6 GB
  • / (/dev/sda1): 45 GB total, 8.9 GB used (21% used), 36 GB available
  • /boot (/dev/sda13): 989 MB total, 196 MB used
  • /boot/efi (/dev/sda15): 105 MB total, 6.3 MB used

1.4 Operating System & Kernel

  • Distribution: Ubuntu 26.04.1 LTS (Resolute Raccoon)
  • Kernel: Linux 7.0.0-1013-oracle (SMP PREEMPT x86_64)
  • Uptime: ~17 hours 31 minutes

2. Running Applications & Workloads

2.1 Docker Containers

Container Name Image Status Port Mapping Storage Mount
new-api calciumion/new-api:latest Up 17+ hours 127.0.0.1:3000 -> 3000/tcp /opt/new-api/data → /data
  • Application Details: New-API (an AI API aggregator and billing dashboard), configured for domain https://aaa.asdkl.sbs with timezone America/Indiana/Indianapolis.

2.2 Web Server / Reverse Proxy (Nginx)

Nginx is running as a reverse proxy with Let's Encrypt SSL/TLS certificates configured: * aaa.asdkl.sbs: * Port 80 → Redirects (HTTP 301) to HTTPS * Port 443 (SSL) → Reverse proxies to http://127.0.0.1:3000 (new-api Docker container) with WebSocket support. * asdkl.sbs: * Port 80 → Redirects to HTTPS * Port 443 (SSL) → Returns plaintext status response: "asdkl.sbs - Oracle server online"


3. Network & Listening Ports

3.1 Exposed Public Ports

  • TCP 22: OpenSSH (sshd)
  • TCP 80: Nginx (HTTP Redirect to HTTPS)
  • TCP 443: Nginx (HTTPS SSL/TLS reverse proxy)

3.2 Internal / Localhost Services

  • TCP 127.0.0.1:3000: Docker proxy for new-api
  • TCP/UDP 127.0.0.53:53 / 127.0.0.54:53: systemd-resolved (DNS stub resolver)
  • UDP 127.0.0.1:323: chronyd (NTP clock synchronization)
  • TCP/UDP 111: rpcbind

3.3 Firewall (iptables)

  • INPUT rules:
  • Accept TCP 80, 443, and 22.
  • Accept established & related connections, ICMP, and loopback traffic.
  • Default policy rejects other unsolicited inbound connections.

4. Key System Daemons

  • Oracle Cloud Agent (snap.oracle-cloud-agent): OCI telemetry, monitoring (gomon), and run-command plugin.
  • Containerd & Docker (docker.service, containerd.service): Container runtime hosting new-api.
  • System Logging & Time: rsyslog, systemd-journald, chronyd.
  • Hardware & Storage Managers: multipathd, iscsid, udisks2.

AI Assistant System Context & Guidelines

1. Hardware Constraints (NON-NEGOTIABLE)

  • Machine: Oracle Cloud Infrastructure VM.Standard.E2.1.Micro (AMD EPYC 7551, 1 OCPU / 2 vCPUs).
  • RAM: ~951 MiB Total + 1.0 GiB Swap.
  • NEVER propose or deploy heavy containerized web panels, Node.js dashboards, or multi-container stacks (e.g., Cockpit, Portainer, Uptime Kuma, Linkwarden). These trigger OOM kills and crash the instance due to extreme swap latency.

2. Server Purpose & Architecture: Fleet Command Center

This server functions as a central command center to manage, update, inspect, and SSH into the user's other remote servers and VMs with zero idle memory consumption.

  • Core Tool: /usr/local/bin/fleet (and symlinked /usr/local/bin/c), written in Python 3 using only standard libraries.
  • Source Code: /home/ubuntu/fleet.py
  • Inventory File: /home/ubuntu/.fleet/inventory.json
  • SSH Keypair: /home/ubuntu/.ssh/id_ed25519 & /home/ubuntu/.ssh/id_ed25519.pub
  • SSH Multiplexing: Configured in ~/.ssh/config using ControlMaster and sockets in ~/.ssh/sockets/.
  • Master Documentation: Read /home/ubuntu/FLEET_MASTER_GUIDE.md for full architecture and command reference.

3. Key User Workflows

  • Interactive Server Login: c or fleet (launches fzf interactive server selector).
  • Direct Login: c <name> or fleet ssh <name>
  • Fleet Updates: fup or fleet update [name|all] (executes apt update && apt upgrade -y in parallel).
  • Command Execution: frun "<cmd>" or fleet run "<cmd>" (runs shell commands across all servers in parallel).
  • Live Fleet Health: fs or fleet status (probes uptime, load, RAM, and disk space across all servers).
  • Add Server: fleet add <name> <ip> [user] [port] [group]
  • Show Public Key: fleet pubkey