The developers of the SUSE project made known, through a publication, the first prototype of ALP (Adaptive Linux Platform), positioned as a continuation of the development of the distribution SUSE Linux Enterprise.
The key difference of the new system is the division of the central framework of the distribution into two parts: A stripped-down "host operating system" to run on top of hardware and an application support layer focused on running in containers and virtual machines.
The idea is to develop in the "host operating system" the minimum environment necessary to support and manage the hardware, and run all applications and components of user space not in a mixed environment, but in separate containers or virtual machines that run on top of the "host operating system" and isolated from each other. This organization will allow users to focus on the abstract applications and workflows of the low-level hardware and system environment.
The idea behind ALP is to allow users to focus on their workloads while abstracting away from the hardware and application layer. Using virtual machines and container technologies, the Adaptable Linux Platform enables workloads to be independent of code flow.
The first prototype of the ALP is called “Les Droites”.
The product SLEMicro, based on the developments of the MicroOS project, is used as the basis for the "host operating system«. For centralized management, Salt (pre-installed) and Ansible (optional) configuration management systems are offered, while Podman and K3s (Kubernetes) are available to run isolated containers. Containerized system components include yast2, podman, k3s, cockpit, GDM (GNOME Display Manager), and KVM.
From the characteristics of the system environment, is mentioned andl default use of disk encryption (FDE, Full Disk Encryption) with the ability to store keys in the TPM, in addition to that partition root is mounted read-only and does not change during operation.
The environment uses the atomic update installation mechanism, unlike the ostree and snap based atomic updates used in Fedora and Ubuntu, ALP uses a normal package manager and snapshot mechanism on the Btrfs file system instead of creating separate atomic images and implement additional delivery infrastructure.
Regarding the part of the basic concepts of ALP, the following are mentioned:
- Minimization of user intervention (zero-touch): involves the automation of the main maintenance, deployment and configuration processes.
- Automatic maintenance of security and search for the system up to date (self-update): provides a configurable mode for automatic update installation (for example, you can enable automatic installation of only fixes for critical vulnerabilities, or revert to manual confirmation of update installation). Live patching is supported to update the Linux kernel without rebooting or suspending work.
- Automatic application of optimizations (self-tuning) and maintenance of system survivability (self-healing): the system captures the latest stable state and after applying updates or changing the configuration in case of detection of anomalies, problems or behavior violations, it is automatically transferred to the previous state using Btrfs snapshots.
- Multiversion software stack: isolate components in containers allows you to use different versions of tools and applications at the same time. For example, you can run applications that depend on different versions of Python, Java, and Node.js by separating incompatible dependencies. Base dependencies come in the form of BCI (Base Container Images) sets. The user can create, update, and remove software stacks without affecting other environments.
Unlike SUSE Linux Enterprise, ALP development is initially carried out through an open development process, in which intermediate builds and test results are publicly available to all, allowing interested parties to make a monitor work in progress and participate in development.
Finally, if you are interested in knowing more about it, you can consult the details In the following link,