scholarly journals CHFS: Parallel Consistent Hashing File System for Node-local Persistent Memory

2022 ◽  
Author(s):  
Osamu Tatebe ◽  
Kazuki Obata ◽  
Kohei Hiraga ◽  
Hiroki Ohtsuji
2021 ◽  
Vol 17 (3) ◽  
pp. 1-25
Author(s):  
Bohong Zhu ◽  
Youmin Chen ◽  
Qing Wang ◽  
Youyou Lu ◽  
Jiwu Shu

Non-volatile memory and remote direct memory access (RDMA) provide extremely high performance in storage and network hardware. However, existing distributed file systems strictly isolate file system and network layers, and the heavy layered software designs leave high-speed hardware under-exploited. In this article, we propose an RDMA-enabled distributed persistent memory file system, Octopus + , to redesign file system internal mechanisms by closely coupling non-volatile memory and RDMA features. For data operations, Octopus + directly accesses a shared persistent memory pool to reduce memory copying overhead, and actively fetches and pushes data all in clients to rebalance the load between the server and network. For metadata operations, Octopus + introduces self-identified remote procedure calls for immediate notification between file systems and networking, and an efficient distributed transaction mechanism for consistency. Octopus + is enabled with replication feature to provide better availability. Evaluations on Intel Optane DC Persistent Memory Modules show that Octopus + achieves nearly the raw bandwidth for large I/Os and orders of magnitude better performance than existing distributed file systems.


Author(s):  
Jianquan Zhang ◽  
Dan Feng ◽  
Jingning Liu ◽  
Lei Yan ◽  
Zheng Zhang

2021 ◽  
Author(s):  
Yang Yang ◽  
Qiang Cao ◽  
Jie Yao ◽  
Yuanyuan Dong ◽  
Weikang Kong

2019 ◽  
Vol 78 (11) ◽  
pp. 973-985 ◽  
Author(s):  
K.Yu. Shekhanin ◽  
A.O. Kolhatin ◽  
E.E. Demenko ◽  
A. A. Kuznetsov
Keyword(s):  

Author(s):  
Mustapha Mohammed Baua'a

The I\O file system Read\Write operations are considered the most significant characteristics. Where, many researchers focus on their works on how to decrease the response time of I\O file system read\write operations. However, most articles concentrate on how to read\write content of the file in parallelism manner. Here in this paper, the author considers the parallelizing Read\Write whole file bytes not only its contents. A case study has been applied in order to make the idea more clear. It talks about two techniques of uploading\downloading files via Web Service. The first one is a traditional way where the files uploaded and downloaded serially. While the second one is uploaded\ downloaded files using Java thread in order to simulate parallelism technique. Java Netbeans 8.0.2 have been used as a programming environment to implement the Download\Upload files through Web Services. Validation results are also presented via using Mat-lab platform as benchmarks. The visualized figures of validation results are clearly clarifying that the second technique shows better response time in comparison to the traditional way.


Sign in / Sign up

Export Citation Format

Share Document