ScholarWorks@성균관대학교

초록

Conventional block interface-based SSDs, performing page-level write operations with out-of-place write policy, lead to issues such as Garbage Collection and Write Amplification. Zoned Namespace (ZNS) SSDs mitigate these issues by segmenting storage into multiple zones and enforcing sequential writes within each zone. Meanwhile, data compression offers effective solutions to increasing storage demands by minimizing data footprint, particularly in high-redundancy environments. This paper addresses the challenges in logical-to-physical page mapping caused by variability in compressed data sizes, analyzes page-level compression characteristics, and discusses strategies for improving compression efficiency by leveraging ZNS SSD features. Using various file types and compression algorithms, the study examines the consistency of compressed data sizes across most file types at the page level. Furthermore, it shows that with a compression ratio of approximately 2.0, up to 86% bandwidth improvement can be achieved in ZNS SSDs.

키워드