Filters








10,728 Hits in 3.8 sec

Improving Disk Throughput in Data-Intensive Servers

E.V. Carrera, R. Bianchini
10th International Symposium on High Performance Computer Architecture (HPCA'04)  
Low disk throughput is one of the main impediments to improving the performance of data-intensive servers.  ...  Our detailed simulations of real server workloads show that FOR and HDC can increase disk throughput by up to 34% and 24%, respectively, in comparison to conventional disk controller cache management techniques  ...  Despite these techniques and optimizations, low disk throughput is still a serious problem for data-intensive servers, such as Web proxies, email and news servers, multimedia servers, and database servers  ... 
doi:10.1109/hpca.2004.10023 dblp:conf/hpca/CarreraB04 fatcat:pqrmlp45yzekfl2sxtmteti42q

Improving disk throughput in data-intensive servers

Enrique V. Carrera, Ricardo Bianchini
2002
Low disk throughput is one of the main impediments to improving the performance of data-intensive servers.  ...  Our detailed simulations of real server workloads show that FOR and HDC can increase disk throughput by up to 34% and 24%, respectively, in comparison to conventional disk controller cache management techniques  ...  Despite these techniques and optimizations, low disk throughput is still a serious problem for data-intensive servers, such as Web proxies, email and news servers, multimedia servers, and database servers  ... 
doi:10.7282/t3-3vnp-wj84 fatcat:fxnlwswjxfdhrae5uvkecikshe

Performance Evaluations of Distributed File Systems for Scientific Big Data in FUSE Environment

Jun-Yeong Lee, Moon-Hyun Kim, Syed Asif Raza Raza Shah, Sang-Un Ahn, Heejun Yoon, Seo-Young Noh
2021 Electronics  
Data are important and ever growing in data-intensive scientific environments.  ...  improvement.  ...  Acknowledgments: The authors would like to extend their sincere thanks to the Global Science Experimental Data Hub Center (GSDC) at the Korea Institute of Science Technology Information (KISTI) for their  ... 
doi:10.3390/electronics10121471 fatcat:bpukewy7ircc5nqrskn64vygga

Data-Intensive Workload Consolidation for the Hadoop Distributed File System

Reza Moraveji, Javid Taheri, Mohammad Reza, Nikzad Babaii Rizvandi, Albert Y. Zomaya
2012 2012 ACM/IEEE 13th International Conference on Grid Computing  
This paper highlights a few challenges of workload consolidation for Hadoop as one of the current state-of-the-art data-intensive cluster computing system.  ...  Through a systematic step-by-step procedure, we investigate challenges for efficient server consolidation in Hadoop environments.  ...  The objective of our study in this paper is to experimentally investigate how to load shared resources of a cluster of servers with data-intensive applications so that their throughput degradation never  ... 
doi:10.1109/grid.2012.25 dblp:conf/grid/MoravejiTRRZ12 fatcat:qgpbwvcii5gldlrohtzjylrhym

IOrchestrator: Improving the Performance of Multi-node I/O Systems via Inter-Server Coordination

Xuechen Zhang, Kei Davis, Song Jiang
2010 2010 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis  
A cluster of data servers and a parallel file system are often used to provide high-throughput I/O service to parallel programs running on a compute cluster.  ...  In this paper we propose a scheme, IOrchestrator, to improve I/O performance of multi-node storage systems by orchestrating I/O services among programs when such inter-data-server coordination is dynamically  ...  This work was also funded in part by the Accelerated Strategic Computing program of the Department of Energy.  ... 
doi:10.1109/sc.2010.30 dblp:conf/sc/ZhangDJ10 fatcat:z4kvpcwzqfg4bcd76fegush65a

Opportunistic Data-driven Execution of Parallel Programs for Efficient I/O Services

Xuechen Zhang, Kei Davis, Song Jiang
2012 2012 IEEE 26th International Parallel and Distributed Processing Symposium  
We propose a data-driven program execution mode in which process scheduling and request issuance are coordinated to facilitate effective I/O scheduling for high disk efficiency.  ...  This data-driven execution mode is enabled when I/O is detected to be the bottleneck, otherwise the program runs in the normal computation-driven mode.  ...  This work was also funded in part by the Accelerated Strategic Computing program of the Department of Energy.  ... 
doi:10.1109/ipdps.2012.39 dblp:conf/ipps/ZhangDJ12 fatcat:jqgnjx3t4vhgzgemhq2xopd54y

A Load-Aware Data Placement Policy on Cluster File System [chapter]

Yu Wang, Jing Xing, Jin Xiong, Dan Meng
2011 Lecture Notes in Computer Science  
In this paper, we present a load-aware data placement policy that will distribute data across the storage servers based on the load of each server and automatically migrate data from heavily-loaded servers  ...  In a large-scale cluster system with many applications running on it, cluster-wide I/O access workload disparity and disk saturation on only some storage servers have been the severe performance bottleneck  ...  In addition to this, LADP-with migration can fully utilize the system aggregate disk bandwidth and improve throughput by data migration.  ... 
doi:10.1007/978-3-642-24403-2_2 fatcat:6ihjfvgawzb3rocxa62z6pyp7e

Exploiting redundancy to boost performance in a RAID-10 style cluster-based file system

Yifeng Zhu, Hong Jiang, Xiao Qin, Dan Feng, David R. Swanson
2006 Cluster Computing  
While aggregating the throughput of existing disks on cluster nodes is a cost-effective approach to alleviate the I/O bottleneck in cluster computing, this approach suffers from potential performance degradations  ...  The duplication of modified data to the mirroring nodes is performed asynchronously in the background.  ...  throughput, implying that when there are communication-intensive applications running on the server nodes, the bottleneck of I/O operations could potentially shift from disks to their TCP/IP stacks.  ... 
doi:10.1007/s10586-006-0011-6 fatcat:ck24a5zpl5gwtjgtfbstxvvska

Active disks for large-scale data processing

E. Riedel, C. Faloutsos, G.A. Gibson, D. Nagle
2001 Computer  
Even in CPU-bound tasks, active disks show linear or near linear improvement with increasing numbers of disks, whereas the traditional server's throughput flatlines in both tests.  ...  Active disks provide a means for accelerating an existing database system by moving data-intensive processing to the disks and off-loading the server CPU.  ... 
doi:10.1109/2.928624 fatcat:4qr6m3f6efglfparnxjjcu3i2m

Improve Throughput of Storage Cluster Interconnected with a TCP/IP Network Using Intelligent Server Grouping [chapter]

Xuechen Zhang, Guiquan Liu, Song Jiang
2010 Lecture Notes in Computer Science  
always results in increased I/O throughput.  ...  Our experimental evaluation shows that SSG can improve I/O throughput by 22.1% on average.  ...  In this way, the long-distance movements of disk heads can be reduced and disk throughput can be improved.  ... 
doi:10.1007/978-3-642-15672-4_31 fatcat:26woxh2tczhmxkbltzkccb6gde

InterFS

Peng Wang, Le Cao, Chunbo Lai, Leqi Zou, Guangyu Sun, Jason Cong
2015 Proceedings of the 6th Asia-Pacific Workshop on Systems - APSys '15  
Therefore, it can be interplanted with other resource-intensive services without interfering with them, and amply fulfill the storage requirements of small-scale applications in the data center.  ...  Resource under-utilization is a common problem in modern data centers.  ...  The imbalance between disk throughput and storage capacity commonly occurs in various data centers. Recently, several approaches have been proposed to improve utilization of computing resources (e.g.  ... 
doi:10.1145/2797022.2797036 dblp:conf/apsys/WangCLZSC15 fatcat:5m75ai2jpbaltantmo7hncr6jy

Atlas: Baidu's key-value storage system for cloud data

Chunbo Lai, Song Jiang, Liqiong Yang, Shiding Lin, Guangyu Sun, Zhenyu Hou, Can Cui, Jason Cong
2015 2015 31st Symposium on Mass Storage Systems and Technologies (MSST)  
All these operations make disk bandwidth available for storing data lower, limiting Atlas's potential improvement on throughput.  ...  Third, to handle the unbalanced use of server resources between metadata service and data service (the former is CPU/Memory/network intensive while the latter is disk I/O intensive), Atlas co-locates the  ... 
doi:10.1109/msst.2015.7208288 dblp:conf/mss/LaiJYLSHCC15 fatcat:2qp7s2lap5hrteyd35ga23r2ea

CTFS: a new lightweight, cooperative temporary file system for cluster-based Web servers

Jun Wang
2003 Proceedings IEEE International Conference on Cluster Computing CLUSTR-03  
Comprehensive tracedriven simulation experiments show that, CTFS achieves up to a 37% better entire system throughput and reduces up to 47% total disk I/O latency than those in asynchronous FFS for a 64  ...  Previous studies showed that I/O could become a major performance bottleneck in cluster-based Web servers.  ...  But disk I/O could easily become a major performance bottleneck for the increasing I/O-intensive workloads in cluster-based Web server.  ... 
doi:10.1109/clustr.2003.1253330 dblp:conf/cluster/Wang03 fatcat:cqkif6f4x5htngx3worvzuyafe

Fan-speed-aware scheduling of data intensive jobs

Christine S. Chan, Yanqin Jin, Yen-Kuan Wu, Kenny Gross, Kalyan Vaidyanathan, Tajana `imuni Rosing
2012 Proceedings of the 2012 ACM/IEEE international symposium on Low power electronics and design - ISLPED '12  
Our measurements show that vibrations induced by fans in high-end servers and its rack neighbors cause a dramatic drop in hard disk bandwidth, resulting in a corresponding decrease in application performance  ...  In this paper we quantify the performance and energy cost effects of the fan vibrations and propose a disk performance aware thermal, energy and cooling technique.  ...  In E6 system JETCX improves energy savings by 1.8x over JETC, and 4.5x as compared to DTM. JETCX gives the best IO throughput for data-intensive jobs at minimal cost to batch job performance.  ... 
doi:10.1145/2333660.2333753 dblp:conf/islped/ChanJWGVR12 fatcat:qajnfjzlk5aepbjexarfsytxlm

Fine-grained device management in an interactive media server

R. Rangaswami, Z. Dimitrijevic, E. Chang, S.-H.G. Chan
2003 IEEE transactions on multimedia  
In this regard, we propose a fine-grained device management strategy consisting of three complementary components: disk profiler, data placement, and IO scheduler.  ...  Through quantitative analysis and experiments, we show that these fine-grained strategies considerably improve device throughput under various workload scenarios.  ...  Zoning placement improves throughput for read-intensive loads by as much as 65%.  ... 
doi:10.1109/tmm.2003.814722 fatcat:3md4vh3lkrfw5hichrf2ejnozq
« Previous Showing results 1 — 15 out of 10,728 results