We're sorry but this page doesn't work properly without JavaScript enabled. Please enable it to continue.
Feedback

Optimized and reproducible HPC Software deployment

Formal Metadata

Title
Optimized and reproducible HPC Software deployment
Subtitle
with free software and GNU Guix
Title of Series
Number of Parts
611
Author
License
CC Attribution 2.0 Belgium:
You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor.
Identifiers
Publisher
Release Date
Language
Production Year2017

Content Metadata

Subject Area
Genre
Abstract
HPC computing environment and especially super computing has its bag ofchallenges when it comes to software deployment. As scientists we often do notget root access which means that we either depend on what software isavailable or we build software in a dedicated directory using tools such asBrew, Conda or even from source. Unfortunately these solutions depend onalready installed tools from an underlying distribution, often proprietary ordated compilers, and, for example, modules. Any binary that gets produced,therefore, tends to be totally unique, both in the generated binary and itsset of dependencies. This is bad, both for trouble shooting and for pursuingreproducible science. With GNU Guix we are developing a fully reproduciblefree software stack including all dependencies targeting cross-builds forheterogeneous computing environments, such as the Beacon Intel PHI-basedsupercomputer, without having a requirement for containers. In this talk Iwill describe the challenges we are overcoming and what it will mean for theHPC community to have architecture optimized and fully reproducible buildtargets that can be deployed in a non-root, non-container environment.