A Distributed Multi-agent Market Place for HPC Compute Cycle Resource Trading

TL;DR

This paper introduces a distributed multi-agent marketplace platform that enables users to buy access to HPC resources, addressing usability and security challenges in distributed computing environments.

Contribution

It presents a novel resource trading platform implemented as a multi-agent system for flexible and efficient scheduling across distributed HPC resources.

Findings

Demonstrates effective resource scheduling in a distributed environment

Shows improved usability and security in HPC resource access

Provides a scalable multi-agent marketplace architecture

Abstract

Computer simulation is finding a role in an increasing number of scientific disciplines, concomitant with the rise in available computing power. Realizing this inevitably requires access to computational power beyond the desktop, making use of clusters, supercomputers, data repositories, networks and distributed aggregations of these resources. Accessing one such resource entails a number of usability and security problems; when multiple geographically distributed resources are involved, the difficulty is compounded. This presents the user with the problem of how to gain access to suitable resources to run their workloads as they need them. In this paper we present our solutions to this problem, a resource trading platform that allows users to purchase access to resources within a distributed e-infrastructure. We present the implementation of this Resource Allocation Market Place as a distributed multi-agent system, and show how it provides a highly flexible, efficient tool to schedule workflows across high performance computing resources.