Resource-Management Study in HPC Runtime-Stacking Context

• Arthur Loussert LaBRI / Univ. Bordeaux / CEA / DAM / DIF
• Benoît Welterlen Bull/Atos SAS
• Patrick Carribault CEA / DAM / DIF
• Julien Jaeger CEA / DAM / DIF
• Marc Pérache CEA / DAM / DIF
• Raymond Namyst LaBRI / Univ. Bordeaux

Resumo

With the advent of multicore and manycore processors as building blocks of HPC supercomputers, many applications shift from relying solely on a distributed programming model (e.g., MPI) to mixing distributed and shared-memory models (e.g., MPI+OpenMP), to better exploit shared-memory communications and reduce the overall memory footprint. One side effect of this programming approach is runtime stacking: mixing multiple models involve various runtime libraries to be alive at the same time and to share the underlying computing resources. This paper explores different configurations where this stacking may appear and introduces algorithms to detect the misuse of compute resources when running a hybrid parallel application. We have implemented our algorithms inside a dynamic tool that monitors applications and outputs resource usage to the user. We validated this tool on applications from CORAL benchmarks. This leads to relevant information which can be used to improve runtime placement, and to an average overhead lower than 1% of total execution time.

Palavras-chave: Runtime, Stacking, Computational modeling, Libraries, Programming, Instruction sets, Runtime library, HPC, Parallel Programming, MPI, OpenMP