The restricted max-min fair allocation problem seeks an allocation of resources to players that maximizes the minimum total value obtained by any player. It is NP-hard to approximate the problem to a ratio less than 2...
详细信息
In this paper we show the first working implementation of an adaptive, measurement-based scheduling algorithm called delay-bounded hybrid proportional delay (DBHPD) in a FreeBSD based ALTQ prototype router. We describ...
详细信息
ISBN:
(纸本)159593197X
In this paper we show the first working implementation of an adaptive, measurement-based scheduling algorithm called delay-bounded hybrid proportional delay (DBHPD) in a FreeBSD based ALTQ prototype router. We describe how we have implemented DBHPD and discuss what kind of difficulties were encountered in the implementation. We present measurement results of the DBHPD implementation with FTP, HTTP, Video Streaming and VoIP traffic in underload, overload and heavy overload conditions. We show that the algorithm operates well according to the theoretical model and preserves the desired delay-bound as well as the delay ratios between the classes. We also compare DBHPD to an existing Class-Based-Queueing (CBQ) algorithm that is widely used both in research and in the industry. We show that DBHPD is able to achieve at least as good link utilization as CBQ and in addition results in much better and predictable differentiation in terms of delays and more controlled packet losses. Copyright 2005 ACM.
This paper proposes a novel optimization technique for MPI persistent communication, that utilizes multiple RDMA engines to carry out low latency communication. Because the interconnects used in modern supercomputers ...
详细信息
ISBN:
(纸本)9788461651337
This paper proposes a novel optimization technique for MPI persistent communication, that utilizes multiple RDMA engines to carry out low latency communication. Because the interconnects used in modern supercomputers have multiple RDMA engines, the multiple communication requests specified by a persistent communication invocation can be scheduled onto the RDMA engines in an optimal way and thus result in better communication performance. Such a scheduling algorithm is not only a packing problem, but also avoids interconnect resource contentions as much as possible. The scheduling algorithm proposed in this paper balances load of RDMA engines and mitigates network link contentions in case of neighbor communication patterns such like in stencil computation. The proposed scheduling algorithm is implemented in Open MPI of K computer using RDMA functions provided as Fujitsu MPI extensions. A typical 2D stencil computation of a climate simulation code is used as a benchmark program. The experimental result shows that a factor of two speedup of communication time is achieved.
This paper presents techniques for the formal specification and efficient incremental implementation of spreadsheet-like tools. The spreadsheets are specified by strong attribute grammars. In this style of attribute g...
详细信息
ISBN:
(纸本)3540201025
This paper presents techniques for the formal specification and efficient incremental implementation of spreadsheet-like tools. The spreadsheets are specified by strong attribute grammars. In this style of attribute grammar programming every single inductive computation is expressed within the attribute grammar formalism. Well-known attribute grammar techniques are used to reason about such grammars. For example, ordered scheduling algorithms can be used to statically guarantee termination of the attribute grammars and to derive efficient implementations. A strong attribute grammar for a spreadsheet is defined and the first incremental results are presented.
this paper, we consider a scheduling algorithm being applied with multi-level security that allows two or more hierarchical classification levels of information to be processed simultaneously. There are various load s...
详细信息
ISBN:
(纸本)9781604234961
this paper, we consider a scheduling algorithm being applied with multi-level security that allows two or more hierarchical classification levels of information to be processed simultaneously. There are various load scheduling algorithms pre-built into the Linux Virtual Server system that have been tested and proven effective for distributing the load among the real servers. While these algorithms may work effectively, there is no current scheduling algorithm that considers a multi-level security protocol to determine which clients have access rights among the servers.
When a train deviates from its planned operational schedule due to unexpected events causing delays, it becomes necessary to reschedule the train in order to obtain a revised operation plan and minimize the delays. Ho...
详细信息
As the source of the content in P2P systems, the main objective of the seed is to help deliver the content to all peers as soon as possible and to ensure the quality of service of the peers as well as to stimulate the...
详细信息
In this paper, we investigate the throughput maximizing problem for a classic three-node relay communication system with a receive-and-forward relay node, which is able to harvest energy from the nature using a rechar...
详细信息
In this paper, we investigate the throughput maximizing problem for a classic three-node relay communication system with a receive-and-forward relay node, which is able to harvest energy from the nature using a rechargeable battery. Besides of the energy consumption for reforwarding the received data, the energy for receiving data at the relay node is also considered. The objective is to maximize the relaying throughput, i.e., receive-and-forward data amount with the energy causality constraints. To this end, we optimize the constituent power allocation for receiving and transmitting under a deterministic energy harvesting (EH) model that the energy arrival time and the energy harvested amount are known prior to relaying. Finally, an algorithm to find the optimal power policy within the constraint of transmission deadline is presented.
The efficient management of resources after a disaster, must take place within a short time and efficiently. Therefore, resource scheduling protocol which shares resources among the victims, respecting time constraint...
详细信息
The efficient management of resources after a disaster, must take place within a short time and efficiently. Therefore, resource scheduling protocol which shares resources among the victims, respecting time constraints, is decisive. In disaster scenarios, communication infrastructure is usually damaged and a commonly used solution for ensuring connectivity between victims and providers is Mobile Ad Hoc Network (MANET). Fire brigade and ambulances might be insufficient if the number of victims is high. Hence, in order to provide an efficient resource scheduling approach that cares about the energy consumption, the number of requesters attended and the processing time, we present a scheduling Resource protocol. Thus, to reduce the provider dislocation time and increase the number of victims attended the protocol was modeled based on Genetic algorithms. Results show that this approach maintains the trade-off between the number of victims attended, therefore, minimize the energy consumption.
In recent years, cloud computing has been extensively used in various fields, where task scheduling plays a vital role for task and resource management. Traditionally, all cores in one system-on-chip (SoC) are treated...
详细信息
ISBN:
(数字)9781665422321
ISBN:
(纸本)9781665446662
In recent years, cloud computing has been extensively used in various fields, where task scheduling plays a vital role for task and resource management. Traditionally, all cores in one system-on-chip (SoC) are treated equally for cloud task scheduling. However, the advent of *** architecture enables heterogeneous computing in a SoC, thus providing the opportunity for task scheduling at SoC level. In this paper, we propose a task scheduling algorithm that leverages performance features of cores on *** architecture. We first characterize the performance of different cores in detail. Then We adopt a decision tree model to predict the execution times of tasks on a single core. Finally we design a task scheduling strategy to speed up the execution speed of big data tasks by comprehensively considering the program characteristics, resource requirements and computing performance on big and LITTLE cores. Experimental results show that our task scheduling algorithm makes more effective use of the computing capabilities of heterogeneous cores, and reduces the running time of big data tasks by 17% on average.
暂无评论