DocumentCode :
180740
Title :
SSD-optimized workload placement with adaptive learning and classification in HPC environments
Author :
Lipeng Wan ; Zheng Lu ; Qing Cao ; Feiyi Wang ; Oral, Sarp ; Settlemyer, Brad
Author_Institution :
Dept. of Electr. Eng. & Comput. Sci., Univ. of Tennessee, Knoxville, TN, USA
fYear :
2014
fDate :
2-6 June 2014
Firstpage :
1
Lastpage :
6
Abstract :
In recent years, non-volatile memory devices such as SSD drives have emerged as a viable storage solution due to their increasing capacity and decreasing cost. Due to the unique capability and capacity requirements in large scale HPC (High Performance Computing) storage environment, a hybrid configuration (SSD and HDD) may represent one of the most available and balanced solutions considering the cost and performance. Under this setting, effective data placement as well as movement with controlled overhead become a pressing challenge. In this paper, we propose an integrated object placement and movement framework and adaptive learning algorithms to address these issues. Specifically, we present a method that shuffle data objects across storage tiers to optimize the data access performance. The method also integrates an adaptive learning algorithm where realtime classification is employed to predict the popularity of data object accesses, so that they can be placed on, or migrate between SSD or HDD drives in the most efficient manner. We discuss preliminary results based on this approach using a simulator we developed to show that the proposed methods can dynamically adapt storage placements and access pattern as workloads evolve to achieve the best system level performance such as throughput.
Keywords :
disc drives; hard discs; information retrieval; learning (artificial intelligence); parallel processing; pattern classification; random-access storage; storage management; HDD drives; HPC environments; SSD drives; SSD-optimized workload placement; adaptive learning algorithms; data access performance optimization; data object accesses; high performance computing storage environment; integrated object movement framework; integrated object placement framework; nonvolatile memory devices; Data models; Drives; Engines; History; Markov processes; Predictive models; Throughput;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Mass Storage Systems and Technologies (MSST), 2014 30th Symposium on
Conference_Location :
Santa Clara, CA
Type :
conf
DOI :
10.1109/MSST.2014.6855552
Filename :
6855552
Link To Document :
بازگشت