Software / roles and alternatives
Some of these programs are part of the experience described in the field notes; others are examples worth understanding, not claims that every one is installed or currently active here. This is deliberately not a complete inventory of a private server.
For each, the more interesting question is why somebody would choose it over a simpler arrangement. Adding software creates another update schedule, another place to store data and another thing to diagnose when it stops. The useful tools earn that work back.
Separates larger workloads
Proxmox runs separate virtual computers and lightweight containers on one physical machine. It helps when several services need different maintenance schedules or an easy way to back up and restore a whole environment. Installing everything directly on one computer is simpler if you only run a few things. Proxmox adds an administrative layer, and several isolated services can still wait on the same unavailable NAS.
Packages smaller services
Docker groups an application with the environment it needs and describes how it connects to stored data and other services. That can make rebuilding easier than repeating a long manual installation. A normal installation may be easier for one program you rarely change. Docker is useful when you run several small applications, but it does not back up their important data automatically or turn an unclear configuration into a clear one.
Optional Docker interface
Portainer is one way to inspect and manage Docker through a visual interface. It can help someone who wants to see containers and their status without remembering every command. Docker's own command-line tools and configuration files remain a perfectly reasonable alternative, especially for a smaller setup. Portainer is discussed here as an option, not as a claim about a specific installation in this home.
Manages shared storage
DSM is the operating system on a Synology NAS. It manages disks, shared folders, health reports and jobs such as backups or snapshots. It appeals to a home user who wants storage administration in one interface rather than building a storage server from separate parts. A self-built server offers more hardware freedom and can cost less, but also leaves more assembly and maintenance to its owner. The storage software and an optional network adapter are separate decisions; my adapter trouble should not be turned into a verdict on every NAS.
Everyday media playback
Plex organizes a personal media library and offers apps for common viewing devices. I keep it as the everyday option because the household knows the apps. That convenience is a practical feature: software everyone can operate does more useful work than a theoretically perfect interface nobody wants to open. Some features and account behaviour differ from alternatives, so the right choice depends on how much you value app availability, local control and a familiar viewing experience.
Another way to play a library
Jellyfin is a self-hosted media platform with a different approach to accounts, control and available apps. It is useful as a comparison when checking whether a playback problem follows a file or a particular application. Someone who prioritizes local control may choose it as their main platform. Running two media platforms at once also means maintaining two sets of software, so the comparison is only worthwhile when the second platform actually answers a question or serves a different need.
Managed network tools
UniFi can bring routing, switching, wireless equipment and security notices into one management view. That becomes useful when a failure involves several wired and wireless devices and you need to see what changed. A basic household router remains a better fit when it already provides good coverage and there is no need for separate network rules or detailed reporting. The common dashboard is convenient, but buying into one product family is a commitment and adds settings that still need somebody to understand them.
Example of an app-specific VPN
WireGuard is a way to make a protected connection; Gluetun can manage that connection for a selected application. The reason to consider app-specific routing is that other devices can keep their ordinary connection. A whole-device VPN is simpler when everything on that device needs the same route. Either way, a connected icon does not prove what happens during failure. This is an architectural example, not a report that I tested a particular protected application here.
Examples of monitoring tools
Uptime Kuma can check whether a page or service answers. Beszel can show processor, memory and storage activity on a monitored computer. Use a simple reachability check when you need to know that something is online; add resource measurements when you need to understand why it is slow. Neither automatically knows whether a visitor can complete the useful task. These are examples of tools for different questions, not a claim about this site's live monitoring arrangement.
Optional library organization
Kometa can maintain collections and artwork rules for a media library. It is useful when a collection is large enough that manual labels become repetitive. For a small library, the media server's built-in organization may already be plenty. Automation deserves a careful preview because one incorrect rule can rearrange many items at once. This is an example of a possible tool, not an inventory of everything running in my home.
How software gets added
I try to add software because a clear need exists: monitor an important service, organize a library, isolate a workload or make recovery more repeatable. Installing a program merely because another homelab screenshot contains it creates maintenance without necessarily creating value.
Before keeping something, I want to know what it stores, how it is backed up, what network access it needs and what will break if it stops. If those questions are difficult to answer, the software may still be useful, but it has not yet become boring enough for permanent employment.
Updates
Updates matter for security, compatibility and fixes. They can also change behaviour. Important services are backed up before major upgrades, and I prefer to read the release information before volunteering the household for an unplanned test program.
Not every update needs to be installed in the first hour. Delaying indefinitely is not a strategy either. The practical middle is to understand the risk, keep a rollback and avoid scheduling major changes immediately before anybody intends to watch something.
When software leaves
A service that duplicates another tool, creates repeated false alarms or demands more care than the problem it solves should be reconsidered. The time already spent configuring it is not a reason to spend more.
Removal includes checking what data, accounts, network rules and scheduled work the tool leaves behind. Deleting the visible box while retaining every exception it created is not simplification. It is archaeology with a shorter menu.