Distributed filesystem metadata updates are typically synchronous. This creates inherent challenges for access efficiency, load balancing, and directory contention, especially under dynamic and skewed workloads. This paper argues that synchronous updates are overly conservative. We propose SwitchFS with asynchronous metadata updates that allow operations to return early and defer directory updates until reads, both hiding latency and amortizing overhead. The key challenge lies in efficiently maintaining the synchronous POSIX semantics of metadata updates. To address this, SwitchFS is co-designed with a programmable switch, leveraging the limited on-switch resources to track directory states with negligible overhead. This allows SwitchFS to aggregate and apply delayed updates efficiently, using batching and consolidation before directory reads. Evaluation shows that SwitchFS achieves up to 13.34× and 3.85× higher throughput, and 61.6% and 57.3% lower latency than two state-of-the-art distributed filesystems, Emulated-InfiniFS and Emulated-CFS, respectively, under skewed workloads. For real-world workloads, SwitchFS improves end-to-end throughput by 21.1×, 1.1×, and 0.3× over CephFS, Emulated-InfiniFS, and Emulated-CFS, respectively.

The Enhanced Read-Only File System (EROFS) has become a popular choice for read-only file systems in a variety of settings, such as mobile devices and cloud services. As EROFS is increasingly applied to diverse contexts, the time-consuming process of creating EROFS images has emerged as a critical factor, often representing a bottleneck in application deployment. However, the process of parallelizing the creation of these images is fraught with challenges, owing to the constraints of fixed-size output compression, the requirement of reproducibility, and the added complexity of de-duplication features. In this paper, we introduce Parer, a parallelized version of mkfs.erofs designed to ensure reproducibility. Parer utilizes a global order-preserving mechanism to assure consistent and reproducible outcomes and employs epoch-based de-duplication to mitigate dependencies during compression. Our evaluation indicates that Parer can reduce the time of creating an EROFS image by up to 79%, while only increasing the image’s size by 3.42% on average. This work has gained recognition from the community and is being synchronized upstream.