What are the benefits and challenges of running docker on Linux versus other platforms?

When oversimplified, every OS has 2 layers: Kernel and Application. Application Layer is where all the runtime environments and codebase resides, whereas Kernel layer is what communicates with HW and helps execute the logic. Docker exists to provide an isolated Application Layer over a Linux Kernel ONLY. So when Docker is run on Linux, it simply uses the host kernel (i.e. Linux). But when Docker is run on any other OS, then a VM with Linux Kernel is setup on top of host. - Benefits of running Docker on Linux over other OS: system overhead, HW utilization, overall latency is lower. - Challenges of running Docker on Linux over other OS: familiarity with command line is required, GUI is usually not present to help navigate.

Docker is primarily a wrapper around Linux kernel features (namespaces, cgroups, capabilities, ...), so If you're using a different host OS, you're going to need a Linux kernel running somewhere anyway. So using it directly on Linux makes the most sense. However, it's not an uncommon use case to run Linux virtualised just for docker, but with the extra layers of complexity, you might have a harder time debugging, should things go south.

Software complexity is indeed the root of all evil. It takes far more skill to design a simple system than cobble together a bunch of garbage software and cross your fingers that it works. Complexity is also a massive security issue.

I belive, Pros and cons are there Pros: - Efficiency: Docker on Linux is resource-efficient, leveraging the host OS kernel. - Native Integration: Linux provides native support, ensuring seamless integration. - Ecosystem: A vast Linux ecosystem enhances Docker with diverse tools. Cons: - Kernel Compatibility: Docker's functionality may vary with kernel versions. - Dependency Issues: Specific Linux distributions may pose compatibility challenges. - Limited GUI Support: Some Linux environments have limited graphical interface support.

I'd argue that a basic familiarity with Linux is a must-have for anyone working in software. Even if you happen to work in .net on Windows, you're still going to find it useful to be familiar with Linux, given that most of the software on earth runs on it. Docker, leveraging the Linux kernel, is part of the Linux ecosystem, so go ahead, and at least learn the basics.

Before starting with Docker, it is better to get familiar with general navigation (cd, ls etc.), file management, package installation, SSH, FTP/SFTP, mount points (/usr, /var, etc.) within Linux environments using command line instead of GUI. A good way to start would be to just install Ubuntu on your laptop/desktop and learn as you go. You can also run a linux machine over any cloud provider and start using it, that works as well. Most importantly, make sure to use command line even if the task is doable by GUI if you really want to learn. As you progress, depending on the use case and with enough internet browsing you'll find all the commands needed.

Someone who does not have enough familiarity with Unix like operating systems to have mastery over a commandline interface should not be doing development or systems administration work at all. This is not a learning curve problem. It is a technical fluency issue. There are far FAR more Unix like systems out there powering our technology that there are graphical systems; systems that are specifically designed to be used by people who do not understand how an operating system functions. Graphical systems are not "simplifying" anything; they are contributing to technical illiteracy. We have enough problems in tech and we don't need that. We need more people who know what they are doing and understand what's happening under the hood.

En el uso de Docker hay una serie de consideraciones importantes respecto a la seguridad: 1. Actualización del servicio de Docker 2. Revisión de vulnerabilidades y actualización de las imágenes. Recomendable usar Trivy. 3. Usar contenedores rootless, limitar permisos del contenedor, implementar espacio de nombre de usuario. 4. Disponer de Registry propio para nuestras imágenes y firmarlas. Recomendable el uso de Harbor. 5. Segmentar la red y comunicación entre los contenedores. 6. Supervisar los contenedores, para detectar comportamientos anómalos (accesos no autorizados, actividades maliciosas, intento de montaje de unidades, etc) 7. No almacenar passwords, claves de API o tokens en texto claro, se recomienda usar "secrets"

This is dead wrong. If you want professional support for Linux, then get professional support. The number of massive corporations who provide professional support is staggering, including VMWare, Redhat, IBM, Oracle, Amazon, any cloud provider, and Microsoft. This is not 1993. If you are running a commercial software system you are doing it on Linux, not Windows. Keep in mind as well that Mac OSX is also a Unix under the hood (BSD based) and all of the same skills apply. This advice is absolutely completely and totally inaccurate. Not surprising an AI came up with it though considering the amount of technical misinformation floating around out there in the general public.