Libra: A Space-Efficient, High-Performance Inline Deduplication for Emerging Hybrid Storage System

Modern hybrid storage systems typically use SSD as the cache layer between memory and primary storage, and adopts inline deduplication to eliminate redundant I/Os and storage, thus delivering better performance and cache endurance. These days, the hybrid storage system is evolved by employing emerging storage mediums as the cache device, aiming at faster access speed and higher endurance. To feed the increasing demand for designing deduplication for emerging cache devices, we propose Libra. Libra leverages the memory and cache cooperated index structure to achieve dual positive benefits of space efficiency and performance. The selective deduplication and historical deduplication metadata recording are also proposed to meet the demands of adopting emerging cache devices. Our experimental results shows that Libra can improve the average bandwidth to $1.13 \times$ and $1.97 \times$ compared to two state-of-the-art deduplication schemes, with much less memory and cache space costs. Also, at most, Libra even improves cache hit rate to $1.53 \times$ and saves 66.9% of cache data space, with the aid of inline deduplication.