DTN Hardware Selection Overview

A DTN server is made of several subsystems. Each needs to be optimized for a DTN workflow. These subsystem include:

The typical engineering trade-offs between cost, redundancy, performance, and so forth apply when deciding on what hardware to use for a DTN node (e.g. redundant power supplies, SATA, SAS, or NVMe backplanes, cost/performance for RAID controllers, quality of NICs). It is recommended to use a system that can be expanded to meet future storage requirements. For the DTN recommendations on this site, we are assuming sequential read-mostly workflows, and a host that is dedicated to data transfer, and not doing data analysis/manipulation.

WARNING: We include some specific hardware examples on these pages, but as hardware changes quickly, it is very difficult to keep this material up to date. Please always check around to see if there are newer, better, cheaper options.

IMPORTANT:  Buying the right hardware is not enough. Extensive tuning is needed for a well optimized DTN.

Here is an Reference DTN Implementation that was the basis for the ESnet Testbed nodes.