Parallax - A New Operating System Prototype Demonstrating Service Scaling and Service Self-Repair in Multi-core Servers

Abstract

This paper describes a prototype demonstrating a new operating system (called Parallax) for the new generation many-core servers. The operating system is designed to address scaling, resource monitoring, dynamic configuration, and self-repair of many-core chip based servers to support distributed computational tasks. The operating system is implemented in assembler language for efficiency and supports C/C++ programming interfaces for high-level programming. The Parallax operating system is based on the DIME network computing model supporting distributed network-centric computing architecture. A parallel signaling network overlay over a network of von Neumann stored program control (SPC) computing nodes is utilized to implement dynamic fault, configuration, accounting, performance, and security management of both the nodes and the network based on business priorities, workload variations and latency constraints.