The Experts below are selected from a list of 87 Experts worldwide ranked by ideXlab platform
Junjie Wang - One of the best experts on this subject based on the ideXlab platform.
-
DASFAA Workshops - h-Buffer: an adaptive buffer management scheme for flash-based storage devices
Database Systems for Advanced Applications, 2012Co-Authors: Rui Wang, Lihua Yue, Peiquan Jin, Junjie WangAbstract:Due to the limitations of flash memory, such as asymmetric I/O latencies and not-in-place update, there are two kinds of buffer replacement algorithms: page-clustered Policy and group-clustered Policy. That the former one organizes pages at page-level makes it easy to deal with hot pages, but shows a bad performance when the buffer size is large enough. The latter one organizes pages at group-level, which usually ignores the read request from the host as the RAM size inside SSDs (Solid State Disks) is limited. However, as the read/write latency for flash memory is about 1:10, and most of desk and server application programs are read-intensive, applying a small portion of buffer space for some hot clean pages will benefit most. In this paper, we propose such a buffer management scheme called h-Buffer with three lists. Applying less than 7.125% of the buffer size for clean pages, h-Buffer considers both the write and read requests by the adoption of a replacement Policy, a Write-Back Policy and a HL (hot list) compensating Policy. Unlike certain existing algorithms, it does not only consider the recency and frequency of page references, but also interacts with the buffer capacities and FTL timely. Experiment results show that the erase count, write count, read count and run time of h-Buffer decrease 50% over traditional algorithms on average.
Bin Cheng - One of the best experts on this subject based on the ideXlab platform.
-
A Machine Learning Based Write Policy for SSD Cache in Cloud Block Storage
2020 Design Automation & Test in Europe Conference & Exhibition (DATE), 2020Co-Authors: Yu Zhang, Ke Zhou, Ping Huang, Hua Wang, Yangtao Wang, Bin ChengAbstract:Nowadays, SSD cache plays an important role in cloud storage systems. The associated write Policy, which enforces an admission control Policy regarding filling data into the cache, has a significant impact on the performance of the cache system and the amount of write traffic to SSD caches. Based on our analysis on a typical cloud block storage system, approximately 47.09% writes are write-only, i.e., writes to the blocks which have not been read during a certain time window. Naively writing the write-only data to the SSD cache unnecessarily introduces a large number of harmful writes to the SSD cache without any contribution to cache performance. On the other hand, it is a challenging task to identify and filter out those write-only data in a real-time manner, especially in a cloud environment running changing and diverse workloads.In this paper, to alleviate the above cache problem, we propose an ML-WP, Machine Learning Based Write Policy, which reduces write traffic to SSDs by avoiding writing write-only data. The main challenge in this approach is to identify write-only data in a real-time manner. To realize ML-WP and achieve accurate write-only data identification, we use machine learning methods to classify data into two groups (i.e., write-only and normal data). Based on this classification, the write-only data is directly written to backend storage without being cached. Experimental results show that, compared with the industry widely deployed Write-Back Policy, ML-WP decreases write traffic to SSD cache by 41.52%, while improving the hit ratio by 2.61% and reducing the average read latency by 37.52%.
-
DATE - A Machine Learning Based Write Policy for SSD Cache in Cloud Block Storage
2020 Design Automation & Test in Europe Conference & Exhibition (DATE), 2020Co-Authors: Yu Zhang, Ke Zhou, Ping Huang, Hua Wang, Yangtao Wang, Hu Jianying, Bin ChengAbstract:Nowadays, SSD cache plays an important role in cloud storage systems. The associated write Policy, which enforces an admission control Policy regarding filling data into the cache, has a significant impact on the performance of the cache system and the amount of write traffic to SSD caches. Based on our analysis on a typical cloud block storage system, approximately 47.09% writes are write-only, i.e., writes to the blocks which have not been read during a certain time window. Naively writing the write-only data to the SSD cache unnecessarily introduces a large number of harmful writes to the SSD cache without any contribution to cache performance. On the other hand, it is a challenging task to identify and filter out those write-only data in a real-time manner, especially in a cloud environment running changing and diverse workloads.In this paper, to alleviate the above cache problem, we propose an ML-WP, Machine Learning Based Write Policy, which reduces write traffic to SSDs by avoiding writing write-only data. The main challenge in this approach is to identify write-only data in a real-time manner. To realize ML-WP and achieve accurate write-only data identification, we use machine learning methods to classify data into two groups (i.e., write-only and normal data). Based on this classification, the write-only data is directly written to backend storage without being cached. Experimental results show that, compared with the industry widely deployed Write-Back Policy, ML-WP decreases write traffic to SSD cache by 41.52%, while improving the hit ratio by 2.61% and reducing the average read latency by 37.52%.
Li-pin Chang - One of the best experts on this subject based on the ideXlab platform.
-
Plugging Versus Logging: Adaptive Buffer Management for Hybrid-Mapping SSDs
ACM Transactions on Embedded Computing Systems, 2015Co-Authors: Li-pin ChangAbstract:A promising technique to improve the write performance of solid-state disks (SSDs) is to use a disk write buffer. The goals of a write buffer is not only to reduce the write traffic to the flash chips but also to convert host write patterns into long and sequential write bursts. This study proposes a new buffer design consisting of a replacement Policy and a Write-Back Policy. The buffer monitors how the host workload stresses the flash translation layer upon garbage collection. This is used to dynamically adjust its replacement and Write-Back strategies for a good balance between write sequentiality and write randomness. When the garbage collection overhead is low, the write buffer favors high write sequentiality over low write randomness. When the flash translation layer observes a high overhead of garbage collection, the write buffer favors low write randomness over high write sequentiality. The proposed buffer design outperformed existing approaches by up to 20p under various workloads and flash translation algorithms, as will be shown in experiment results.
-
DAC - Plugging versus logging: a new approach to write buffer management for solid-state disks
Proceedings of the 48th Design Automation Conference on - DAC '11, 2011Co-Authors: Li-pin ChangAbstract:Using device write buffers is a promising technique to improve the write performance of solid-state disks. The write buffer not only reduces the write traffic to the flash but also produces large and sequential write bursts to the underlying flash translation layer. This study proposes a new buffer design consisting of a replacement Policy and a Write-Back Policy. This buffer monitors how the host workload stresses the flash translation layer upon garbage collection, and dynamically adjusts its replacement and Write-Back strategies for a good balance between write sequentiality and traffic reduction. Experimental results show that the proposed buffer design outperformed existing approaches by up to 20% under various workloads and flash translation algorithms.
Wang Xiang-na - One of the best experts on this subject based on the ideXlab platform.
-
An Implementation of RAID5 Based Disk Cache
Computer Engineering, 2004Co-Authors: Wang Xiang-naAbstract:This article introduces an implementation of RAID5 based disk cache. In the process of implementing the disk array cache, we used some mature techniques, set-associated map, LRU replace Policy, and so on. By using Write-Back Policy, we improved the speed of writing to disk and reduced redundant disk write operations. In addition, by locking the parity group, we effectively prevented data inconsistency caused by simultaneous degrade of multi blocks within the same parity group.
Rui Wang - One of the best experts on this subject based on the ideXlab platform.
-
DASFAA Workshops - h-Buffer: an adaptive buffer management scheme for flash-based storage devices
Database Systems for Advanced Applications, 2012Co-Authors: Rui Wang, Lihua Yue, Peiquan Jin, Junjie WangAbstract:Due to the limitations of flash memory, such as asymmetric I/O latencies and not-in-place update, there are two kinds of buffer replacement algorithms: page-clustered Policy and group-clustered Policy. That the former one organizes pages at page-level makes it easy to deal with hot pages, but shows a bad performance when the buffer size is large enough. The latter one organizes pages at group-level, which usually ignores the read request from the host as the RAM size inside SSDs (Solid State Disks) is limited. However, as the read/write latency for flash memory is about 1:10, and most of desk and server application programs are read-intensive, applying a small portion of buffer space for some hot clean pages will benefit most. In this paper, we propose such a buffer management scheme called h-Buffer with three lists. Applying less than 7.125% of the buffer size for clean pages, h-Buffer considers both the write and read requests by the adoption of a replacement Policy, a Write-Back Policy and a HL (hot list) compensating Policy. Unlike certain existing algorithms, it does not only consider the recency and frequency of page references, but also interacts with the buffer capacities and FTL timely. Experiment results show that the erase count, write count, read count and run time of h-Buffer decrease 50% over traditional algorithms on average.
