Components and Interfaces of a Process Management System for Parallel Programs

TL;DR

This paper introduces MPD, a scalable process management system designed for parallel programs like MPI, enabling faster startup, improved runtime control, and better integration with parallel debugging and utilities.

Contribution

The paper presents MPD, a scalable, efficient process management system for parallel jobs, with a flexible interface separating process management from parallel libraries.

Findings

Faster startup times for large parallel jobs.

Enhanced control over stdio for parallel processes.

Support for parallel debugging and utilities.

Abstract

Parallel jobs are different from sequential jobs and require a different type of process management. We present here a process management system for parallel programs such as those written using MPI. A primary goal of the system, which we call MPD (for multipurpose daemon), is to be scalable. By this we mean that startup of interactive parallel jobs comprising thousands of processes is quick, that signals can be quickly delivered to processes, and that stdin, stdout, and stderr are managed intuitively. Our primary target is parallel machines made up of clusters of SMPs, but the system is also useful in more tightly integrated environments. We describe how MPD enables much faster startup and better runtime management of parallel jobs. We show how close control of stdio can support the easy implementation of a number of convenient system utilities, even a parallel debugger. We describe a simple but general interface that can be used to separate any process manager from a parallel library, which we use to keep MPD separate from MPICH.