Application-Grained Block I/O Analysis for Edge Computational Intelligent

Block I/O workload is widely used in studies on storage performance. However, few studies have focused on the workload characteristics of edge intelligent computing. A discrepancy exists between the workload and the storage, resulting in performance loss. Most studies have focused on the overall block I/O workload characteristics or the performance of the storage; however, a fine-grained analysis is required to understand the workload more comprehensively. In this paper, we propose the concept of an application-grained block I/O analysis for edge computational intelligent. Firstly describe the five steps of the application-grained block I/O analysis, i.e., trace collection, application isolation, application weighting, application trace analysis, and results. Subsequently, discuss the key points of the application-grained block I/O analysis. An application-grained block I/O analysis is performed using five off-the-shelf trace analysis tools and four evaluation parameters.

[1]  Sai Narasimhamurthy,et al.  Characterizing Deep-Learning I/O Workloads in TensorFlow , 2018, 2018 IEEE/ACM 3rd International Workshop on Parallel Data Storage & Data Intensive Scalable Computing Systems (PDSW-DISCS).

[2]  Anil Kashyap,et al.  Workload Characterization for Enterprise Disk Drives , 2018, ACM Trans. Storage.

[3]  Reza Salkhordeh,et al.  ReCA: An Efficient Reconfigurable Cache Architecture for Storage Systems with Online Workload Characterization , 2018, IEEE Transactions on Parallel and Distributed Systems.

[4]  David Hung-Chang Du,et al.  Hot Data Identification with Multiple Bloom Filters: Block-Level Decision vs I/O Request-Level Decision , 2018, Journal of Computer Science and Technology.

[5]  Guangjie Han,et al.  Coordinate Memory Deduplication and Partition for Improving Performance in Cloud Computing , 2019, IEEE Transactions on Cloud Computing.

[6]  Gala Yadgar,et al.  Avoiding the Streetlight Effect: I/O Workload Analysis with SSDs in Mind , 2016, HotStorage.

[7]  Yifeng Zhu,et al.  Design and Implementation of a Hybrid Shingled Write Disk System , 2016, IEEE Transactions on Parallel and Distributed Systems.

[8]  Feng Chen,et al.  Understanding storage I/O behaviors of mobile applications , 2016, 2016 32nd Symposium on Mass Storage Systems and Technologies (MSST).

[9]  Guangjie Han,et al.  A Maximum Cache Value Policy in Hybrid Memory-Based Edge Computing for Mobile Devices , 2019, IEEE Internet of Things Journal.

[10]  Nong Xiao,et al.  Red: An efficient replacement algorithm based on REsident Distance for exclusive storage caches , 2010, 2010 IEEE 26th Symposium on Mass Storage Systems and Technologies (MSST).

[11]  Alma Riska,et al.  Disk Drive Level Workload Characterization , 2006, USENIX Annual Technical Conference, General Track.

[12]  Gregory R. Ganger,et al.  Track-Aligned Extents: Matching Access Patterns to Disk Drive Characteristics , 2002, FAST.

[13]  Dandan Wang,et al.  Larger , Cheaper , but Faster : SSD-SMR Hybrid Storage Boosted by a New SMR-oriented Cache Framework , 2017 .

[14]  Peter Desnoyers,et al.  Analytic Models of SSD Write Performance , 2014, TOS.

[15]  Qi Zhang,et al.  Characterization of storage workload traces from production Windows Servers , 2008, 2008 IEEE International Symposium on Workload Characterization.

[16]  Tim Brecht,et al.  Disk Prefetching Mechanisms for Increasing HTTP Streaming Video Server Throughput , 2018, ACM Trans. Model. Perform. Evaluation Comput. Syst..

[17]  Guangjie Han,et al.  Hybrid-LRU Caching for Optimizing Data Storage and Retrieval in Edge Computing-Based Wearable Sensors , 2019, IEEE Internet of Things Journal.

[18]  Antony I. T. Rowstron,et al.  Write off-loading: Practical power management for enterprise storage , 2008, TOS.

[19]  Qingyue Liu.  Ouroboros Wear-leveling: A Two-level Hierarchical Wear-leveling Model for NVRAM , 2017 .

[20]  Hai Jin,et al.  LaLDPC: Latency-aware LDPC for Read Performance Improvement of Solid State Drives , 2017 .