Staying Secure & Up to Date – Part 2: Know What You Run (Inventory)
“I don’t even know what runs on my server” – you are in good company
Before you can secure anything, you need to know what you operate. Sounds trivial, but it’s the #1 reason security measures miss the target: you patch the one thing you know and overlook the three services that have been running unnoticed for years.
So let’s do an inventory. The goal is a single list that answers at a glance:
In the next parts, this list will become your personal checklist – and you rarely need to change it.
Step 1: Collect your devices
Take a sheet of paper or a Markdown file and write down every device that offers any kind of service – in short: everything that is reachable from the outside:
• Your desktop / laptop
• A NAS or small home server
• A mini-PC with virtualization (e.g. Proxmox) – and inside it all virtual machines and
container systems
• A Raspberry Pi (Pi-hole, WireGuard …)
• And don’t forget: rented small servers on the internet (VPS) and even your router
A simplified example from a real environment:
| Device | Role | OS | |----------------|-------------------------------------|------------------------| | Desktop | daily work, Docker for a tool | KDE neon (Ubuntu 24.04)| | Mini-PC “H1” | hypervisor for VMs & containers | Proxmox VE 9 | | VM “proxy” | web entry point, WAF/proxy | Debian 13 | | VM “cloud” | Nextcloud server | Debian 13 | | Server “A” | monitoring, DB replica | Ubuntu 24.04 | | Server “B” | websites, mail, Paperless, vault | Ubuntu 24.04 |
Step 2: Two essential values per device
To make sense of updates you need two values per device – you can check them right away with the commands below in your terminal:
1. Operating system & version – e.g. “Ubuntu 24.04”, “Debian 13”, “Proxmox 9.2”
2. Kernel version – e.g. 6.8.0-147
Linux:
cat /etc/os-release | grep PRETTY_NAME uname -r
Windows (PowerShell):
systeminfo | findstr /B /C:"OS Name" /C:"OS Version"
Proxmox: pveversion (or the GUI header – it usually shows
pve-manager/9.2.x (running kernel: 6.x)).
Step 3: Running services & containers
On Docker hosts, one command shows all containers: docker ps
Example excerpt from a real inventory (yours will look similar):
NAMES IMAGE STATUS traefik traefik:3.7.14 Up 3 weeks (healthy) paperless-webserver ghcr.io/paperless-ngx/paperless-ngx:3.4.1 Up 4 days (healthy) minio quay.io/minio/aistor/minio:RELEASE.2026-09-19 Up 1 day (healthy) vaultwarden vaultwarden/server:latest Up 3 days (healthy)
Note especially: which service (Traefik = reverse proxy, Paperless = document archive,
MinIO = object storage, Vaultwarden = password server) and which image (with
:latest you know: “gets updated continuously, but needs monitoring” – part 7).
Step 4: Open updates per system
A good moment to capture the update status:
Ubuntu/Debian:
sudo apt update apt list --upgradable
Ubuntu with Pro/ESM (free for personal use on up to 5 machines – part 6):
sudo pro status sudo pro security-status
Proxmox:
sudo apt update apt list --upgradable
Collect all results in a table with a third column: “updates open? when last patched”. This will be the core of your monthly check (part 8).
Pro Corner
• Pin exact versions, not just names: for containers, record the exact image
tag (e.g. ghcr.io/paperless-ngx/paperless-ngx:3.4.1) – :latest is a
moving target, not a version.
• SBOM as a pro inventory: generate a Software Bill of Materials per
image, e.g. with Trivy (trivy image --format cyclonedx <image>).
• Don’t forget kernel coupling: on Proxmox, all LXCs share the host kernel – patching the hypervisor patches several systems at once. Put the hypervisor at the top of your routine.
• Maintain the inventory automatically: a small cron/systemd job that writes
docker ps, pveversion, and apt state into a file (or your notes system) saves
manual bookkeeping – we build that concretely in parts 5 and 8.
Why this matters: you can only manage what you know. The list from part 2 is your starting point for every future scan and routine.
Next part: Part 3 – Your toolbox: scanners, databases & sources.