WSL Containers on Windows 11: A Docker Alternative?

Windows 11 has introduced an exciting feature: WSL Containers (wslc), allowing users to run Linux containers without relying on Docker. This development invites users to assess whether wslc can serve as a viable alternative to Docker Desktop, particularly for those using Windows as their primary operating system. A recent hands-on exploration reveals the workings and capabilities of WSL Containers, highlighting both strengths and weaknesses compared to established container technologies.

What Are WSL Containers?

WSL Containers, also referred to as wslc, are an innovative addition to the Windows Subsystem for Linux (WSL). They differ markedly from Docker; each wslc container operates within its own dedicated virtual machine (VM), while Docker Desktop runs containers inside a shared WSL utility VM. This architecture makes wslc particularly well-suited for development environments, although it raises questions about performance and resource management for long-running services.

Performance Evaluation of WSL Containers

When assessing wslc's performance, initial expectations pointed toward a competitive edge regarding speed and resource efficiency. Testing revealed that while initial reports indicated rapid container startup times for wslc, real-world metrics showed performance equivalence with Docker Desktop. The following figures illuminate the startup times for both container solutions:

While wslc did show slightly faster warm container start times, the differences were minimal and not perceptible in practice. The consistent performance across multiple tests highlights the reliability of wslc when running development workloads.

Memory Management Concerns

A key concern raised during the evaluation was the RAM consumption associated with wslc. Each wslc container initiates its own VM, consuming about 1.5 GB of RAM. In contrast, Docker Desktop remains relatively lightweight, utilizing under 200 MB of memory while idling. However, the real challenge appears when containers do not terminate their associated VMs properly, leading to lingering memory usage even after sessions appear closed. This issue necessitates manual intervention through the command-line interface (CLI), often requiring a command to terminate the entire wslc session.

Exploring GPU Accelerated Workloads

One of the standout capabilities of wslc is its performance with GPU workloads, particularly concerning CUDA-enabled applications. In tests, CUDA 12.4 images seamlessly recognized the system’s RTX 5090 GPU, which demonstrates wslc's efficacy in managing hardware resources efficiently. It is important to note that using GPU acceleration with wslc requires a simple flag setup without the complexities often associated with driver configurations.

Networking Stack Differences

The networking stack in wslc presents some inconsistencies compared to Docker. For instance, all network bindings default to 127.0.0.1, as opposed to the 0.0.0.0 that Docker uses. In testing, this necessitated explicit publishing of the networking address; otherwise, containers would not communicate as expected. Such discrepancies can lead to complications, particularly for users unfamiliar with the required configurations.

Availability of Docker Compose Functionality

Another significant gap in wslc's offerings is the lack of direct support for Docker Compose. As it stands, Microsoft has not integrated this crucial feature, which often leads users back to Docker Desktop for comprehensive container management. However, it is noteworthy that a community-driven project, known as wslc-compose, has emerged to fill this void. This solution introduces a Python shim that allows users to work with Compose files in a manner aligned with wslc requirements, effectively bridging the gap until official support is implemented by Microsoft.

Usability and Visibility Challenges

A notable shortcoming identified in wslc is its visibility. Unlike Docker Desktop, which provides a graphical user interface (GUI) to monitor and manage containers, wslc relies primarily on command-line tools. This can lead to a steeper learning curve for users accustomed to the visual management tools that Docker offers, often requiring more effort to maintain awareness of running services.

Conclusion: Is WSL Containers the Future?

Given the comprehensive analysis, wslc emerges as a promising alternative to Docker Desktop, particularly for developers interested in performance and ease of use for dev containers. With GPU support functioning effectively and a solid foundation for creating containerized applications, wslc is set to become a vital tool. However, visibility issues and the lack of built-in Compose support remain challenges that must be addressed for widespread adoption. Users intrigued by wslc can access it via the Windows Subsystem for Linux and experiment with its features using the pre-release updates, setting the stage for a future where a straightforward Docker alternative is not just viable but competitive.