Impact of Exploiting Load Imbalance on Coscheduling in Workstation Clusters

Grid is an emerging infrastructure which enables effective coordinate access to various distributed computing resources in order to serve the needs of collaborative research and work across the world. Grid resource management is always a key subject in the grid computing. We first analyze the resource management in the grid computing environment, then according to the load imbalance question in the ant colony optimization algorithm, propose an improved algorithm that suits to be used in the grid environment.

Download Full-text

Load imbalance in DASD dynamic reconnection

Information Processing Letters ◽

10.1016/0020-0190(88)90155-x ◽

1988 ◽

Vol 28 (3) ◽

pp. 111-119

Author(s):

Alexandre Brandwajn

Keyword(s):

Load Imbalance

Download Full-text

Low-latency message passing on workstation clusters using SCRAMNet

Proceedings 13th International Parallel Processing Symposium and 10th Symposium on Parallel and Distributed Processing. IPPS/SPDP 1999 ◽

10.1109/ipps.1999.760450 ◽

2003 ◽

Cited By ~ 5

Author(s):

V. Moorthy ◽

M.G. Jacunski ◽

M. Pillai ◽

P.P. Ware ◽

D.K. Panda ◽

...

Keyword(s):

Message Passing ◽

Low Latency ◽

Workstation Clusters

Download Full-text

Scalable mining of maximal quasi-cliques

Proceedings of the VLDB Endowment ◽

10.14778/3436905.3436916 ◽

2020 ◽

Vol 14 (4) ◽

pp. 573-585

Author(s):

Guimu Guo ◽

Da Yan ◽

M. Tamer Özsu ◽

Zhe Jiang ◽

Jalal Khalil

Keyword(s):

Graph Mining ◽

State Of The Art ◽

Minimum Degree ◽

The Other ◽

Dense Subgraph ◽

Load Imbalance ◽

Subgraph Mining ◽

Execution Engine ◽

Clique Algorithm ◽

Dense Subgraph Mining

Given a user-specified minimum degree threshold γ , a γ -quasiclique is a subgraph g = (V g , E g ) where each vertex ν ∈ V g connects to at least γ fraction of the other vertices (i.e., ⌈ γ · (| V g |- 1)⌉ vertices) in g. Quasi-clique is one of the most natural definitions for dense structures useful in finding communities in social networks and discovering significant biomolecule structures and pathways. However, mining maximal quasi-cliques is notoriously expensive. In this paper, we design parallel algorithms for mining maximal quasi-cliques on G-thinker, a distributed graph mining framework that decomposes mining into compute-intensive tasks to fully utilize CPU cores. We found that directly using G-thinker results in the straggler problem due to (i) the drastic load imbalance among different tasks and (ii) the difficulty of predicting the task running time. We address these challenges by redesigning G-thinker's execution engine to prioritize long-running tasks for execution, and by utilizing a novel timeout strategy to effectively decompose long-running tasks to improve load balancing. While this system redesign applies to many other expensive dense subgraph mining problems, this paper verifies the idea by adapting the state-of-the-art quasi-clique algorithm, Quick, to our redesigned G-thinker. Extensive experiments verify that our new solution scales well with the number of CPU cores, achieving 201× runtime speedup when mining a graph with 3.77M vertices and 16.5M edges in a 16-node cluster.

Download Full-text

DATA ALLOCATION STRATEGIES FOR PARALLEL IMAGE PROCESSING ALGORITHMS

International Journal of Pattern Recognition and Artificial Intelligence ◽

10.1142/s0218001495000237 ◽

1995 ◽

Vol 09 (04) ◽

pp. 615-634 ◽

Cited By ~ 3

Author(s):

VIRGINIE MARION-POTY ◽

SERGE MIGUET

Keyword(s):

Parallel Implementation ◽

Image Data ◽

Data Allocation ◽

Load Imbalance ◽

Parallel Image Processing ◽

Parallel Image ◽

Image Processing Algorithms ◽

Processing Algorithms ◽

Recursive Operators ◽

Allocation Strategies

This paper discusses several data allocation strategies used for the parallel implementation of basic imaging operators. It shows that depending on the operator (sequential or parallel, with regular or irregular execution time), the image data must be partitioned in very different manners: The square sub-domains are best adapted for minimizing the communication volume, but rectangles can perform better when we take into account the time for constructing messages. Block allocations are well adapted for inherently parallel operators since they minimize interprocessor interactions, but in the case of recursive operators, they lead to nearly sequential executions. In this framework, we show the usefulness of block-cyclic allocations. Finally, we illustrate the fact that allocating the same amount of image data to each processor can lead to severe load imbalance in the case of some operators with data-dependant execution times.

Download Full-text

THE LOAD BALANCING OF SELF-SIMILAR TRAFFIC IN NETWORK INTRUSION DETECTION SYSTEMS

Cybersecurity Education Science Technique ◽

10.28925/2663-4023.2020.7.1730 ◽

2020 ◽

Vol 3 (7) ◽

pp. 17-30

Author(s):

Tamara Radivilova ◽

Lyudmyla Kirichenko ◽

Maksym Tawalbeh ◽

Petro Zinchenko ◽

Vitalii Bulakh

Keyword(s):

Load Balancing ◽

Intrusion Detection ◽

Intrusion Detection System ◽

Detection System ◽

Intrusion Detection Systems ◽

Deep Packet Inspection ◽

Detection Systems ◽

Network Intrusion ◽

Load Imbalance ◽

Packet Inspection

The problem of load balancing in intrusion detection systems is considered in this paper. The analysis of existing problems of load balancing and modern methods of their solution are carried out. Types of intrusion detection systems and their description are given. A description of the intrusion detection system, its location, and the functioning of its elements in the computer system are provided. Comparative analysis of load balancing methods based on packet inspection and service time calculation is performed. An analysis of the causes of load imbalance in the intrusion detection system elements and the effects of load imbalance is also presented. A model of a network intrusion detection system based on packet signature analysis is presented. This paper describes the multifractal properties of traffic. Based on the analysis of intrusion detection systems, multifractal traffic properties and load balancing problem, the method of balancing is proposed, which is based on the funcsioning of the intrusion detection system elements and analysis of multifractal properties of incoming traffic. The proposed method takes into account the time of deep packet inspection required to compare a packet with signatures, which is calculated based on the calculation of the information flow multifractality degree. Load balancing rules are generated by the estimated average time of deep packet inspection and traffic multifractal parameters. This paper presents the simulation results of the proposed load balancing method compared to the standard method. It is shown that the load balancing method proposed in this paper provides for a uniform load distribution at the intrusion detection system elements. This allows for high speed and accuracy of intrusion detection with high-quality multifractal load balancing.

Download Full-text