David Hung-Chang Du
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
David Hung-Chang Du has a total of 27 co-authors. Here are the top co-authors by number of shared publications:
- Ewa Kusmierek
Poznań Supercomputing and Networking Center
2 shared publications - Fenggang Wu
Meta Platforms, Inc., Sunnyvale, USA
2 shared publications James A. MacDonald
2 shared publications- Jenwei Hsieh2 shared publications
Mengjou Lin
2 shared publications- Dongchul Park
Department of Computer Science and Engineering, University of Minnesota-Twin Cities, Minneapolis, USA
1 shared publication
Latest Publications
- 27 December 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cloud Computing
- Vol. 9 (2) , 627-640
Guaranteed Bang for the Buck: Modeling VDI Applications to Identify Storage Requirements
In the cloud environment, most services are provided by virtual machines (VMs). Identifying storage requirements of VMs is challenging, but it is essential for good user experiences while optimizing use of storage resources. Determining the storage configuration necessary to support and satisfy VMs first requires an accurate description of the VM configurations, and the problem is further exacerbated by the diversity and special characteristics of the VMs. In this paper, we study Virtual Desktop Infrastructure (VDI), a prevalent and complicated VM application, to identify and characterize storage requirements of VMs and determine how to meet such requirements with minimal storage resources and cost. We first create a model to describe the behavior of VDI, and we collect real VDI traces to populate this model. The model allows us to identify the storage requirements of VDI and determine the potential bottlenecks of a given storage configuration. Based on this information, we can tell what capacity and minimum capability a storage configuration needs in order to support and satisfy given VDI configurations. We show that our model can describe more fine-grained VM behavior varying with time and virtual disk types compared with the rules of thumb currently used in industry.
- 27 December 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Cloud Computing
- Vol. 9 (2) , 627-640
Guaranteed Bang for the Buck: Modeling VDI Applications to Identify Storage Requirements
In the cloud environment, most services are provided by virtual machines (VMs). Identifying storage requirements of VMs is challenging, but it is essential for good user experiences while optimizing use of storage resources. Determining the storage configuration necessary to support and satisfy VMs first requires an accurate description of the VM configurations, and the problem is further exacerbated by the diversity and special characteristics of the VMs. In this paper, we study Virtual Desktop Infrastructure (VDI), a prevalent and complicated VM application, to identify and characterize storage requirements of VMs and determine how to meet such requirements with minimal storage resources and cost. We first create a model to describe the behavior of VDI, and we collect real VDI traces to populate this model. The model allows us to identify the storage requirements of VDI and determine the potential bottlenecks of a given storage configuration. Based on this information, we can tell what capacity and minimum capability a storage configuration needs in order to support and satisfy given VDI configurations. We show that our model can describe more fine-grained VM behavior varying with time and virtual disk types compared with the rules of thumb currently used in industry.
- 18 June 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 68 (1) , 111-124
On Improving the Write Responsiveness for Host-Aware SMR Drives
This paper presents a Virtual Persistent Cache design to remedy the long latency behavior and to ultimately improve the write responsiveness of the Host-Aware Shingled Magnetic Recording (HA-SMR) drives. Our design keeps the cost-effective model of the existing HA-SMR drives, but at the same time asks the great help from the host system for adaptively providing some computing and management resources to improve the drive performance when needed. The technical contribution is to trick the HA-SMR drives by smartly reshaping the access patterns to HA-SMR drives, so as to avoid the occurrences of long latencies in most cases and thus to ultimately improve the drive performance and responsiveness. We conduct experiments on real Seagate 8 TB HA-SMR drives to demonstrate the advantages of Virtual Persistent Cache over the real workloads from Microsoft Research Cambridge. The results show that the proposed design can remedy most of the long latencies and improve the drive performance by at least 58.11%, under the evaluated workloads.
- 18 June 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 68 (1) , 111-124
On Improving the Write Responsiveness for Host-Aware SMR Drives
This paper presents a Virtual Persistent Cache design to remedy the long latency behavior and to ultimately improve the write responsiveness of the Host-Aware Shingled Magnetic Recording (HA-SMR) drives. Our design keeps the cost-effective model of the existing HA-SMR drives, but at the same time asks the great help from the host system for adaptively providing some computing and management resources to improve the drive performance when needed. The technical contribution is to trick the HA-SMR drives by smartly reshaping the access patterns to HA-SMR drives, so as to avoid the occurrences of long latencies in most cases and thus to ultimately improve the drive performance and responsiveness. We conduct experiments on real Seagate 8 TB HA-SMR drives to demonstrate the advantages of Virtual Persistent Cache over the real workloads from Microsoft Research Cambridge. The results show that the proposed design can remedy most of the long latencies and improve the drive performance by at least 58.11%, under the evaluated workloads.
- 11 January 2017
- Published by Springer Nature in Journal of Computer Science and Technology
- Vol. 32 (1) , 26-40
A Lookahead Read Cache: Improving Read Performance for Deduplication Backup Storage
- 11 January 2017
- Published by Springer Nature in Journal of Computer Science and Technology
- Vol. 32 (1) , 26-40
A Lookahead Read Cache: Improving Read Performance for Deduplication Backup Storage
- 29 November 2015
- Published by Springer Nature
- p. 339-350
Link Delay Modeling for Two-Way Traffic Road Segment in Vehicular Networks
- 29 November 2015
- Published by Springer Nature
- p. 339-350
Link Delay Modeling for Two-Way Traffic Road Segment in Vehicular Networks
- 23 June 2007
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 36 (3) , 243-266
Proxy-assisted periodic broadcast for video streaming with multiple servers
- 23 June 2007
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 36 (3) , 243-266
Proxy-assisted periodic broadcast for video streaming with multiple servers
- 31 March 2007
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 34 (3) , 267-297
Periodic broadcast with dynamic server selection
- 31 March 2007
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 34 (3) , 267-297
Periodic broadcast with dynamic server selection
- 17 November 2006
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 33 (2) , 131-156
Full-sharing: efficient bandwidth scheduling for video streaming over broadband cable networks (BCNs)
- 17 November 2006
- Published by Springer Nature in Multimedia Tools and Applications
- Vol. 33 (2) , 131-156
Full-sharing: efficient bandwidth scheduling for video streaming over broadband cable networks (BCNs)
- 1 September 2005
- Published by Taylor & Francis in International Journal of Parallel, Emergent and Distributed Systems
- Vol. 20 (3-4) , 221-233
Energy efficient organization of mobile sensor networks
One of the main design issues for a sensor network is conservation of energy available at each sensor node. To increase lifetime of a sensor network, we can organize the sensors into disjoint sets so that only one set is active at a time. The lifetime of the network increases proportionally to the number of disjoint sets. In general, sensors are randomly scattered in the monitored area, thus the number of disjoint sets is significantly smaller than in the ideal case. In this paper, we propose a way to increase the number of disjoint sets using mobile sensors. We organize sensors into disjoint sets using the heuristic proposed by Sljepcevic and Potkonjak, Power efficient organization of wireless sensor networks IEEE International Conference on Communications (ICC) 2001. Then, we rearrange mobile sensors that are not included in any set. The rearrangement process consists of two phases. In the first phase, we identify the fields that are not covered by any of the remaining sensors. Then we identify the locations from which mobile sensors could cover the fields. In the second phase, our proposed heuristic selects sensors to be rearranged and locations where they will move to. This selection is made by considering the coverage before and after a mobile sensor is moved. Our experiments show that we can effectively increase the number of disjoint set with a small number of mobile sensors rearranged.
- 1 September 2005
- Published by Taylor & Francis in International Journal of Parallel, Emergent and Distributed Systems
- Vol. 20 (3-4) , 221-233
Energy efficient organization of mobile sensor networks
One of the main design issues for a sensor network is conservation of energy available at each sensor node. To increase lifetime of a sensor network, we can organize the sensors into disjoint sets so that only one set is active at a time. The lifetime of the network increases proportionally to the number of disjoint sets. In general, sensors are randomly scattered in the monitored area, thus the number of disjoint sets is significantly smaller than in the ideal case. In this paper, we propose a way to increase the number of disjoint sets using mobile sensors. We organize sensors into disjoint sets using the heuristic proposed by Sljepcevic and Potkonjak, Power efficient organization of wireless sensor networks IEEE International Conference on Communications (ICC) 2001. Then, we rearrange mobile sensors that are not included in any set. The rearrangement process consists of two phases. In the first phase, we identify the fields that are not covered by any of the remaining sensors. Then we identify the locations from which mobile sensors could cover the fields. In the second phase, our proposed heuristic selects sensors to be rearranged and locations where they will move to. This selection is made by considering the coverage before and after a mobile sensor is moved. Our experiments show that we can effectively increase the number of disjoint set with a small number of mobile sensors rearranged.
- 31 March 2005
- Published by Elsevier in Journal of Systems and Software
- Vol. 75 (3) , 237-252
Streaming video delivery over Internet with adaptive end-to-end QoS
- 31 March 2005
- Published by Elsevier in Journal of Systems and Software
- Vol. 75 (3) , 237-252
Streaming video delivery over Internet with adaptive end-to-end QoS