FAST (Fast Agent System Timers) is a tool for dynamic performance forecasting in a Grid environment. FAST is composed of several layers and relies on low level software. First, it uses a network and CPU monitoring software to handle dynamically changing resources, like workload or bandwidth. FAST uses the Network Weather Service (NWS), a distributed system that periodically monitors and dynamically forecasts the performance of various network and computational resources. The dynamic data acquisition module of FAST uses and enhances NWS. FAST also includes routines to model the time and space needs for each triplet (problem ; machine ; parameters set ). They are based on benchmarking at installation time on each machine for a representative set of parameters and polynomial data fitting. To store these static data, FAST uses a LDAP tree.


MUMPS (MUltifrontal Massively Parallel Solver) is a parallel software package (Fortran 90, MPI) for the solution of sparse systems of linear equations. Its performance is very competitive and it has a large number of functionalities, including:

  • types of systems: symmetric (positive definite or not), or unsymmetric,
  • various matrix input formats (assembled or elemental, distributed or centralized),
  • threshold partial pivoting for robustness,
  • partial factorization (computation of a Schur complement matrix),
  • real or complex arithmetic,
  • fully asynchronous approach with distributed dynamic scheduling,
  • C or Fortran 90 interface,
  • ...
MUMPS is available free of charge.

Huge problems can now be computed over the Internet thanks to Grid Computing Environments like Globus or Legion. Because most of current applications are numerical, the use of libraries like BLAS, LAPACK, ScaLAPACK or PETSc is mandatory. The integration of such libraries in high level applications using languages like Fortran or C is far from being easy. Moreover, the computational power and memory needs of such applications may of course not be available on every workstation. Thus, the RPC seems to be a good candidate to build Problem Solving Environments on the Grid. Several tools following this approach exist, like Netsolve, NINF, NEOS, or RCS. The aim of the DIET project is to develop a set of tools to build computational servers.


Since the advent of distributed computer systems an active field of research has been the investigation of scheduling strategies for parallel applications. The common approach is to employ scheduling heuristics that approximate an optimal schedule. Unfortunately, it is not possible to obtain analytical results to compare and select appropriate heuristics for a given scheduling problem. One possibility is to conducts large numbers of back-to-back experiments on real platforms. While this is possible on tightly-coupled platforms, it is infeasible on modern distributed platforms such as the Computational Grid as it is labor-intensive and does not enable repeatable results. The solution is to resort to simulations. Simulations not only lead to repeatable results but also make it possible to explore wide ranges of platform and application scenarios. This can be addressed by using a trace-based approach in which the behavior of the platform is recorded via monitoring tools and ``replayed'' in simulation. SimGrid is a conjoint effort with UCSD to provide a framework which enables the simulation of distributed applications in distributed computing environments for the specific purpose of developing and evaluating scheduling algorithms.

Eddy Caron
Last modified: Wed Dec 4 00:38:36 GMT 2002