Installation and Troubleshooting Guide

This guide describes how to install, debug, and isolate the performance and functional problems that are related to GDS and is intended for systems administrators and developers.

This guide describes how to debug and isolate the NVIDIA® Magnum IO GPUDirect® Storage (GDS) related performance and functional problems and is intended for systems administrators and developers.

GDS enables a direct data path for direct memory access (DMA) transfers between GPU memory and storage, which avoids a bounce buffer through the CPU. This direct path increases system bandwidth and decreases the latency and utilization load on the CPU.

Creating this direct path involves distributed file systems such as NFSoRDMA, DDN EXAScaler parallel file system solutions (based on the Lustre file system), Amazon FSx for Lustre, and WekaFS, so the GDS environment is composed of multiple software and hardware components. This guide addresses questions related to the GDS installation and helps you triage functionality and performance issues. For non-GDS issues, contact the respective OEM or file systems vendor to understand and debug the issue.