Characterizing, Modeling and Predicting Dynamic Resource Availability in a Large Scale Multi-purpose Grid

Author(s):  
Farrukh Nadeem ◽  
Radu Prodan ◽  
Thomas Fahringer
Author(s):  
Philip Chan ◽  
David Abramson

Wide-area distributed systems offer new opportunities for executing large-scale scientific applications. On these systems, communication mechanisms have to deal with dynamic resource availability and the potential for resource and network failures. Connectivity losses can affect the execution of workflow applications, which require reliable data transport between components. We present the design and implementation of p-channels, an asynchronous and fault-tolerant pipe mechanism suitable for coupling workflow components. Fault-tolerant communication is made possible by persistence, through adaptive caching of pipe segments while providing direct data streaming. We present the distributed algorithm for implementing: (a) caching of pipe data segments; (b) asynchronous read operation; and (c) communication state transfer to handle dynamic process joins and leaves.


Author(s):  
R. Jeyarani ◽  
N. Nagaveni ◽  
R. Vasanth Ram

Cloud Computing provides dynamic leasing of server capabilities as a scalable, virtualized service to end users. The discussed work focuses on Infrastructure as a Service (IaaS) model where custom Virtual Machines (VM) are launched in appropriate servers available in a data-center. The context of the environment is a large scale, heterogeneous and dynamic resource pool. Nonlinear variation in the availability of processing elements, memory size, storage capacity, and bandwidth causes resource dynamics apart from the sporadic nature of workload. The major challenge is to map a set of VM instances onto a set of servers from a dynamic resource pool so the total incremental power drawn upon the mapping is minimal and does not compromise the performance objectives. This paper proposes a novel Self Adaptive Particle Swarm Optimization (SAPSO) algorithm to solve the intractable nature of the above challenge. The proposed approach promptly detects and efficiently tracks the changing optimum that represents target servers for VM placement. The experimental results of SAPSO was compared with Multi-Strategy Ensemble Particle Swarm Optimization (MEPSO) and the results show that SAPSO outperforms the latter for power aware adaptive VM provisioning in a large scale, heterogeneous and dynamic cloud environment.


2005 ◽  
Vol 62 (4) ◽  
pp. 913-924 ◽  
Author(s):  
Karl M Polivka

I used field observations, assays, and experiments with the euryhaline cottid Cottus aleuticus to evaluate the extent to which average resource availability drives the large-scale distribution of these fish among upstream and estuarine habitats and how interspecific competition from a congener affects its performance in the estuary. Population densities of C. aleuticus were only consistent with resource densities across years during two of five study years, indicating a lack of resource matching at large temporal scales. On shorter temporal scales, fish growth rates that were two to three times higher in the estuary compared with the stream were inconsistent with the predictions of resource matching theory. A manipulation of C. aleuticus density showed that the estuary could support at least twice the number of individuals that occurred there; thus, the profitable estuary is underutilized. Interspecific competition with Cottus asper was partially responsible for this underutilization as indicated by a substantial reduction in growth and condition among C. aleuticus individuals in experimental manipulations that compared intra- and inter-specific effects. Observed spatial overlap between these two cottids combined with the results of the competition experiment suggests that C. aleuticus is more strongly limited in its ability to use estuarine habitats opportunistically by interspecific competition than by intraspecific competition.


2021 ◽  
Vol 7 ◽  
pp. e824
Author(s):  
Yiren Li ◽  
Tieke Li ◽  
Pei Shen ◽  
Liang Hao ◽  
Wenjing Liu ◽  
...  

Microservice-based Web Systems (MWS), which provide a fundamental infrastructure for constructing large-scale cloud-based Web applications, are designed as a set of independent, small and modular microservices implementing individual tasks and communicating with messages. This microservice-based architecture offers great application scalability, but meanwhile incurs complex and reactive autoscaling actions that are performed dynamically and periodically based on current workloads. However, this problem has thus far remained largely unexplored. In this paper, we formulate a problem of Dynamic Resource Scheduling for Microservice-based Web Systems (DRS-MWS) and propose a similarity-based heuristic scheduling algorithm that aims to quickly find viable scheduling schemes by utilizing solutions to similar problems. The performance superiority of the proposed scheduling solution in comparison with three state-of-the-art algorithms is illustrated by experimental results generated through a well-known microservice benchmark on disparate computing nodes in public clouds.


2019 ◽  
Vol 30 (4) ◽  
pp. 814-826 ◽  
Author(s):  
Dazhao Cheng ◽  
Xiaobo Zhou ◽  
Yinggen Xu ◽  
Liu Liu ◽  
Changjun Jiang

2002 ◽  
Vol 71 (6) ◽  
pp. 994-1001 ◽  
Author(s):  
Patrik Byholm ◽  
Esa Ranta ◽  
Veijo Kaitala ◽  
Harto Linden ◽  
Pertti Saurola ◽  
...  

Sign in / Sign up

Export Citation Format

Share Document