Peter J. Keleher

Using content-addressable networks for load balancing in desktop grids

Desktop grids have evolved to combine Peer-to-Peer and Grid computing techniques to improve the robustness, reliability and scalability of job execution infrastructures. However, efficiently matching incoming jobs to available system resources and achieving good load balance in a fully decentralized and heterogeneous computing environment is a challenging problem. In this paper, we extend our prior work with a new decentralized algorithm for maintaining approximate global load information, and a job pushing mechanism that uses the global information to push jobs towards underutilized portions of the system. The resulting system more effectively balances load and improves overall system throughput. Through a comparative analysis of experimental results across different system configurations and job profiles, performed via simulation, we show that our system can reliably execute Grid applications on a distributed set of resources both with low cost and with good load balance.

Resource Discovery Techniques in Distributed Desktop Grid Environments

Desktop grids use opportunistic sharing to exploit large collections of personal computers and workstations across the Internet, achieving tremendous computing power at low cost. Traditional desktop grid systems are typically based on a client-server architecture, which has inherent shortcomings with respect to robustness, reliability and scalability. In this paper, we propose a decentralized, robust, highly available, and scalable infrastructure to match incoming jobs to available resources. Through a comparative analysis on the experimental results obtained via simulation of three different types of matchmaking algorithms under different workload scenarios, we show the trade-offs between effcient matchmaking and good load balancing in a fully decentralized, heterogeneous computational environment.

Creating a Robust Desktop Grid using Peer-to-Peer Services

The goal of the work described in this paper is to design and build a scalable infrastructure for executing grid applications on a widely distributed set of resources. Such grid infrastructure must be decentralized, robust, highly available, and scalable, while efficiently mapping application instances to available resources in the system. However, current desktop grid computing platforms are typically based on a client-server architecture, which has inherent shortcomings with respect to robustness, reliability and scalability. Fortunately, these problems can be addressed through the capabilities promised by new techniques and approaches in peer-to-peer (P2P) systems. By employing P2P services, our system allows users to submit jobs to be run in the system and to run jobs submitted by other users on any resources available in the system, essentially allowing a group of users to form an ad-hoc set of shared resources. The initial target application areas for the desktop grid system are in astronomy and space science simulation and data analysis.

Trade-offs in matching jobs and balancing load for distributed desktop grids

Desktop grids can achieve tremendous computing power at low cost through opportunistic sharing of resources. However, traditional client-server Grid architectures do not deal with all types of failures, and do not always cope well with very dynamic environments. This paper describes the design of a desktop grid implemented over a modified Peer-to-Peer (P2P) architecture. The underlying P2P system is decentralized and inherently adaptable, giving the Grid robustness, scalability, and the ability to cope with dynamic environments, while still efficiently mapping application instances to available resources throughout the system. We use simulation to compare three different types of matching algorithms under differing workloads. Overall, the P2P approach produces significantly lower wait times than prior approaches, while adapting efficiently to the dynamic environment.

Decentralized, accurate, and low-cost network bandwidth prediction

The distributed nature of modern computing makes end-to-end prediction of network bandwidth increasingly important. Our work is inspired by prior work that treats the Internet and bandwidth as an approximate tree metric space. This paper presents a decentralized, accurate, and low cost system that predicts pairwise bandwidth between hosts. We describe an algorithm to construct a distributed tree that embeds bandwidth measurements. The correctness of the algorithm is provable when driven by precise measurements. We then describe three novel heuristics that achieve high accuracy for predicting bandwidth even with imprecise input data. Simulation experiments with a real-world dataset confirm that our approach shows high accuracy with low cost.

Decentralized resource management for multi-core desktop grids

The majority of CPUs now sold contain multiple computing cores. However, current desktop grid computing systems either ignore the multiplicity of cores, or treat them as distinct, independent machines. The latter approach ignores the resource contention present between cores in a single CPU, while the former approach fails to take advantage of significant computing power. We propose a decentralized resource management framework for exploiting multi-core nodes in peer-to-peer grids. We present two new load-balancing schemes that explicitly account for the resource sharing and contention of multiple cores, and propose a simple simulation model that can represent a continuum of resource sharing among cores of a CPU. We use simulation to confirm that our two algorithms match jobs w ith computing nodes efficiently, and balance load during the lifetime of the computing jobs.

Integrating categorical resource types into a P2P desktop grid system

We describe and evaluate a set of protocols that implement a distributed, decentralized desktop grid. Incoming jobs are matched with system nodes through proximity in an N-dimensional resource space. This work improves on prior work by (1) efficiently accommodating node and job characterizations that include both continuous and categorical resource types, and (2) scaling gracefully to large system sizes even with highly non-uniform distributions of job and node types. We use extensive simulation results to show that the resulting system handles both continuous and categorical constraints efficiently, and that the new scalability techniques are effective.

Decentralized dynamic scheduling across heterogeneous multi-core desktop grids

The recent advent of multi-core computing environments increases both the heterogeneity and complexity of managing desktop grid resources, making efficient load balancing challenging even for a centralized manager. Even with good initial job assignments, dynamic scheduling is still needed to adapt to dynamic environments, as well as for applications whose running times are not known a priori. In this paper, we propose new decentralized scheduling schemes that backfill jobs locally and dynamically migrate waiting jobs across nodes to leverage residual resources, while guaranteeing bounded waiting times for all jobs. The methods attempt to maximize total throughput while balancing load across available grid resources. Experimental results via simulation show that our scheduling scheme has performance competitive with an online centralized scheduler.

Searching for Bandwidth-Constrained Clusters

Data-intensive distributed applications can increase their performance by running on a cluster of hosts connected via high-bandwidth interconnections. However, there is no effective method to find such a bandwidth-constrained cluster in a decentralized fashion. Our work is inspired by prior work that treats Internet bandwidth as an approximate tree metric space. This paper presents a decentralized, accurate, and efficient method to find a cluster of Internet hosts, given the desired cluster size and minimum interconnection bandwidth. We describe a centralized polynomial time algorithm for a tree metric space, along with a proof of correctness. We then provide a decentralized version of the algorithm. Simulation experiments with two real-world datasets confirm that our clustering approach achieves high accuracy and scalability. We also discuss the costs of decentralization and how the treeness of the dataset affects clustering accuracy.

Supporting Computing Element Heterogeneity in P2P Grids

We propose resource discovery and load balancing techniques to accommodate computing nodes with many types of computing elements, such as multi-core CPUs and GPUs, in a peer-to-peer desktop grid architecture. Heterogeneous nodes can have multiple types of computing elements, and the performance and characteristics of each computing element can be very different. Our scheme takes into account these diverse aspects of heterogeneous nodes to maximize overall system throughput. However, straightforward methods of handling diverse computing elements that differ on many axes can result in high overheads, both in local state and in communication volume. We describe approaches that minimize messaging costs without sacrificing the failure resilience provided by an underlying peer-to-peer overlay network. Simulation results show that our schemes load balancing performance is comparable to that of a centralized approach, that communication costs are reduced significantly compared to the existing system, and that failure resilience is not compromised.