FTL: A new operating system for clouds

You can now run Linux binaries in containers that are as secure as VMs, with OS features as libraries.

FTL, a new operating system designed for cloud environments, has been released. It operates as a userspace OS, accessible as a library, allowing developers to build and manage OS functionalities as if they were applications. This design simplifies feature addition, debugging, and safe upgrades. FTL's kernel isolates containers with a lightweight, hardware-based isolation mechanism, aiming for security comparable to virtual machines without compromising performance. It supports Linux binaries and offers an alternative to traditional monolithic kernels by combining microkernel flexibility with monolithic kernel simplicity. Key features include improved container security, support for Linux binaries, and the ability to run specialized applications without POSIX abstractions. The project roadmap includes support for Node.js, Go, SMP, container images, and 64-bit Arm architecture by early 2027.

FTL's novel userspace OS architecture, where the OS functions as a library, signals a departure from traditional monolithic kernel designs in cloud computing. By leveraging a minimal kernel interface and hardware-based isolation, FTL aims to bridge the gap between the security of virtual machines and the efficiency of containers. This approach could significantly impact how cloud-native applications are developed and deployed, offering developers greater control and flexibility over their operating environment.

The ability to safely extend OS features, debug, and upgrade within a library-based userspace OS represents a paradigm shift for cloud developers. Compatibility with existing Linux binaries suggests a smooth adoption path for many applications, while the roadmap points towards broader language and architecture support. This development may lead to more specialized, secure, and performant cloud deployments, potentially reducing the overhead associated with traditional VM-based solutions for high-security workloads.