214. Introduction

MPI is primarily concerned with communication rather than process or resource management. However, it is necessary to address these issues to some degree in order to define a useful framework for communication. This chapter presents a set of MPI interfaces that allows for a variety of approaches to process management while placing minimal restrictions on the execution environment.

The MPI model for process creation allows both the creation of an intial set of processes related by their membership in a common MPI_COMM_WORLD and the creation and management of processes after an MPI application has been started. A major impetus for the latter form of process creation comes from the PVM [24] research effort. This work has provided a wealth of experience with process management and resource control that illustrates their benefits and potential pitfalls.

The MPI Forum decided not to address resource control because it was not able to design a portable interface that would be appropriate for the broad spectrum of existing and potential resource and process controllers. Resource control can encompass a wide range of abilities, including adding and deleting nodes from a virtual parallel machine, reserving and scheduling resources, managing compute partitions of an MPP, and returning information about available resources. MPI assumes that resource control is provided externally --- probably by computer vendors, in the case of tightly coupled systems, or by a third party software package when the environment is a cluster of workstations.

The reasons for including process management in MPI are both technical and practical. Important classes of message-passing applications require process control. These include task farms, serial applications with parallel modules, and problems that require a run-time assessment of the number and type of processes that should be started. On the practical side, users of workstation clusters who are migrating from PVM to MPI may be accustomed to using PVM's capabilities for process and resource management. The lack of these features would be a practical stumbling block to migration.

The following goals are central to the design of MPI process management:

• The MPI process model must apply to the vast majority of current parallel environments. These include everything from tightly integrated MPPs to heterogeneous networks of workstations.
• MPI must not take over operating system responsibilities. It should instead provide a clean interface between an application and system software.

• MPI must guarantee communication determinism in the presense of dynamic processes, i.e., dynamic process management must not introduce unavoidable race conditions.

• MPI must not contain features that compromise performance.

Second, MPI maintains a consistent concept of a communicator, regardless of how its members came into existence. A communicator is never changed once created, and it is always created using deterministic collective operations.