File systems unfit as distributed storage backends: lessons from 10 years of Ceph evolution

A Aghayev, S Weil, M Kuchnik, M Nelson… - Proceedings of the 27th …, 2019 - dl.acm.org
For a decade, the Ceph distributed file system followed the conventional wisdom of building
its storage backend on top of local file systems. This is a preferred choice for most distributed …

Finding {Crash-Consistency} Bugs with Bounded {Black-Box} Crash Testing

J Mohan, A Martinez, S Ponnapalli, P Raju… - … USENIX Symposium on …, 2018 - usenix.org
We present a new approach to testing file-system crash consistency: bounded black-box
crash testing (B3). B3 tests the file system in a black-box manner using workloads of file …

MittOS: Supporting millisecond tail tolerance with fast rejecting SLO-aware OS interface

M Hao, H Li, MH Tong, C Pakha, RO Suminto… - Proceedings of the 26th …, 2017 - dl.acm.org
MittOS provides operating system support to cut millisecond-level tail latencies for data-
parallel applications. In MittOS, we advocate a new principle that operating system should …

Towards Better Understanding of Black-box {Auto-Tuning}: A Comparative Analysis for Storage Systems

Z Cao, V Tarasov, S Tiwari, E Zadok - 2018 USENIX Annual Technical …, 2018 - usenix.org
Modern computer systems come with a large number of configurable parameters that control
their behavior. Tuning system parameters can provide significant gains in performance but is …

Carver: Finding important parameters for storage system tuning

Z Cao, G Kuenning, E Zadok - 18th USENIX Conference on File and …, 2020 - usenix.org
Storage systems usually have many parameters that affect their behavior. Tuning those
parameters can provide significant gains in performance. Alas, both manual and automatic …

Is garbage collection overhead gone? case study of F2FS on ZNS SSDs

D Seo, PX Chen, H Li, M Bjørling, N Dutt - … of the 15th ACM Workshop on …, 2023 - dl.acm.org
The sequential write nature of ZNS SSDs makes them very well-suited for log-structured file
systems. The Flash-Friendly File System (F2FS), is one such log-structured file system and …

{MIDAS}: Minimizing Write Amplification in {Log-Structured} Systems through Adaptive Group Number and Size Configuration

S Oh, J Kim, S Han, J Kim, S Lee, SH Noh - 22nd USENIX Conference on …, 2024 - usenix.org
Log-structured systems are widely used in various applications because of its high write
throughput. However, high garbage collection (GC) cost is widely regarded as the primary …

Performance evaluation of host aware shingled magnetic recording (HA-SMR) drives

F Wu, Z Fan, MC Yang, B Zhang, X Ge… - IEEE Transactions on …, 2017 - ieeexplore.ieee.org
Shingled Magnetic Recording (SMR) drives can benefit large-scale storage systems by
reducing the Total Cost of Ownership (TCO) of dealing with explosive data growth. Among …

Building efficient key-value stores via a lightweight compaction tree

T Yao, J Wan, P Huang, X He, F Wu, C Xie - ACM Transactions on …, 2017 - dl.acm.org
Log-Structure Merge tree (LSM-tree) has been one of the mainstream indexes in key-value
systems supporting a variety of write-intensive Internet applications in today's data centers …

Zonetier: A zone-based storage tiering and caching co-design to integrate ssds with smr drives

X Xie, L Xiao, DHC Du - ACM Transactions on Storage (TOS), 2019 - dl.acm.org
Integrating solid-state drives (SSDs) and host-aware shingled magnetic recording (HA-SMR)
drives can potentially build a cost-effective high-performance storage system. However …