Unikernels are specialized, single-address-space op-erating systems (OSes) tailored to specific applications. They offer strong isolation, low memory/disk footprints, and fast startup times-making them well-suited for cloud and serverless computing. However, deploying many of them at scale in cloud environments introduces new challenges. In particular, managing library updates and versioning in statically linked unikernels is difficult due to their tightly coupled structure. Unlike dynamically linked binaries, statically linked unikernels lack built-in versioning mechanisms. Consequently, even minor library changes result in entirely new memory layouts, which can significantly increase memory consumption when multiple instances run concurrently. We present Spacer-A, a framework that improves memory sharing across statically linked unikernels with different library versions. Spacer uses differential analysis and library align-ment to enable page-level sharing via memory deduplication scanners or a custom loader backed by a shared library pool. Our evaluation with Unikraft shows that Spacer reduces memory consumption and boot overhead while maintaining compatibility across versions. The framework integrates into existing unikernel build pipelines with minimal changes and is released as open source.

Unikernels offer strong isolation and performance benefits over traditional virtual machines, but their specialized, statically linked design complicates memory deduplication—particularly in multi-tenant environments. Traditional approaches like Kernel Samepage Merging (KSM) struggle with convergence delays, high CPU usage, and unpredictable memory savings. Dynamic linking improves memory reuse but compromises performance, simplicity, and security. We present Spacer-SLT, a novel load time deduplication mechanism that eliminates the need for memory scanners. Spacer-SLT extracts common libraries into a shared pool and uses a custom Firecracker-based loader to enable instant, deterministic memory deduplication at launch time. Unlike KSM, it introduces no runtime overhead and maintains the benefits of static linking, including performance and reduced attack surface.

Supporting mainstream applications is fundamental for a new OS to have impact. It is generally achieved by developing a layer of compatibility allowing applications developed for a mainstream OS like Linux to run unmodified on the new OS. Building such a layer, as we show, results in large engineering inefficiencies due to the lack of efficient methods to precisely measure the OS features required by a set of applications.

Unikernels are on the rise in the cloud. These lightweight virtual machines (VMs) specialized to a single application offer the same level of isolation as full-blown VMs, while providing performance superior to standard Linux-based VMs or even to containers. However, their inherent specialization renders memory deduplication ineffective, causing unikernels, in practice, to consume more memory than their small memory footprint would suggest. This makes them less advantageous when thousands of SaaS and/or FaaS unikernels instances have to run on the same server.

Unikernels are famous for providing excellent performance in terms of boot times, throughput and memory consumption, to name a few metrics. However, they are infamous for making it hard and extremely time consuming to extract such performance, and for needing significant engineering effort in order to port applications to them. We introduce Unikraft, a novel micro-library OS that (1) fully modularizes OS primitives so that it is easy to customize the unikernel and include only relevant components and (2) exposes a set of composable, performance-oriented APIs in order to make it easy for developers to obtain high performance.