Evaluation of Various Workloads in Filebench Suitable for Phase-change Memory

Lee, Seungyong; Lee, Hyokeun; Lee, Hyuk-Jae; Kim, Hyun

doi:10.5573/ieiespc.2021.10.2.160

Cited by 3 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We use Filebench [16] and RocksDB [52] as two real-world applications to evaluate the efficacy of different FTL designs.…”

Section: Real-world Applicationsmentioning

confidence: 99%

“…By contrast, since LearnedFTL exploits both the spatial locality and the learned model, it achieves 0.3% and 46% CMT hit ratio, 55% and 41% model hit ratio in random reads and sequential reads, respectively. Filebench: Filebench [16] is a highly flexible storage benchmark. We select three workloads that are most widely used in previous studies [7,20,64]: fileserver (write heavy), webserver (read heavy, less random write), and randomread (all random reads).…”

Section: Real-world Applicationsmentioning

confidence: 99%

See 1 more Smart Citation

LearnedFTL: A Learning-based Page-level FTL for Improving Random Reads in Flash-based SSDs

Wang¹,

Lin²,

Wu³

et al. 2023

Preprint

View full text Add to dashboard Cite

We present LearnedFTL, which applies learned indexes to on-demand page-level flash translation layer (FTL) designs to improve the random read performance of flash-based solidstate drives (SSDs). The first of its kind, it minimizes the number of double reads induced by address translation in random read accesses. To apply the learned indexes to address translation, LearnedFTL proposes dynamic piece-wise regression to efficiently build learned indexes. LearnedFTL also exploits the unique feature of page relocation in SSD internal garbage collection (GC), and embeds the learned index training in GC, which can minimize additional delay on normal read and write operations. Additionally, LearnedFTL employs a bitmap prediction filter to guarantee the accuracy of learned indexes' predictions. With these designs, LearnedFTL considerably speeds up address translation while reducing the number of flash read accesses caused by the demand-based page-level FTL. Our benchmark-driven experiments on a FEMU-based prototype show that LearnedFTL reduces the 99th percentile tail latency by 4.8×, on average, compared to the state-of-the-art TPFTL scheme.

show abstract

“…We use Filebench [16] and RocksDB [52] as two real-world applications to evaluate the efficacy of different FTL designs.…”

Section: Real-world Applicationsmentioning

confidence: 99%

Section: Real-world Applicationsmentioning

confidence: 99%

LearnedFTL: A Learning-based Page-level FTL for Improving Random Reads in Flash-based SSDs

Wang¹,

Lin²,

Wu³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Conversely, Linux OS is a standalone operating system that provides the underlying functionality for KVM to operate independently. As a result, it can be claimed that KVM above the primary OS (type 2 hypervisor), employing already developed system functionalities in the absence of its own [13][14].…”

Section: Kvm and Ms Hyper-vmentioning

confidence: 99%

Mathematical Modeling and File System Performance Examination for Type 1 Hypervisor with Full Hardware Virtualization in the Case of KVM and MS Hyper-V

Đorđević¹,

Kraljević²,

Štrbac-Savić³

et al. 2023

Proceedings of the International Scientific Conference - Sinteza 2023

View full text Add to dashboard Cite

The paper analyses the performance comparison of hypervisor type 1 in the case of KVM and MS Hyper-V platforms for virtualization. The characteristics of both tested hypervisors were examined through file system performance. The tests were performed under equal conditions and with an inimitable testing methodology, using the benchmark program -Filebench. The performance of the tested hypervisors was compared taking into account tests performed for a system with one, two and finally three virtual machines in state of operation. Mathematical modelling was done, which is the main contribution of this work, then hypotheses about the expected behaviour were set and confirmed through the obtained results. Ubuntu 22.04 LTS, an OS (operating system) from the Debian distribution family, was execute as the guest OS.

show abstract

Why Address Translation Matter?: Analyzing Page Access Patterns in NAND Flash-Based SSDs

Kim,

Hong

2023

2023 International Technical Conference on Circuits/Systems, Computers, and Communications (ITC-CSCC)

View full text Add to dashboard Cite

Evaluation of Various Workloads in Filebench Suitable for Phase-change Memory

Cited by 3 publications

References 0 publications

LearnedFTL: A Learning-based Page-level FTL for Improving Random Reads in Flash-based SSDs

LearnedFTL: A Learning-based Page-level FTL for Improving Random Reads in Flash-based SSDs

Mathematical Modeling and File System Performance Examination for Type 1 Hypervisor with Full Hardware Virtualization in the Case of KVM and MS Hyper-V

Why Address Translation Matter?: Analyzing Page Access Patterns in NAND Flash-Based SSDs

Contact Info

Product

Resources

About