Workshop 6: HPC Compute Cluster Workshop
LISA: Where systems engineering and operations professionals share real-world knowledge about designing, building, and maintaining the critical systems of our interconnected world.
The LISA conference has long served as the annual vendor-neutral meeting place for the wider system administration community. The LISA14 program recognized the overlap and differences between traditional and modern IT operations and engineering, and developed a highly-curated program around 5 key topics: Systems Engineering, Security, Culture, DevOps, and Monitoring/Metrics. The program included 22 half- and full-day training sessions; 10 workshops; and a conference program consisting of 50 invited talks, panels, refereed paper presentations, and mini-tutorials.
Ballard Room
Clay England, Oak Ridge National Laboratory
Administering a compute cluster in a production environment is a niche area of system administration. In addition to the common issues involved in administering *NIX computers, additional challenges related to cluster management, customer usage, and specialized software present themselves. In this workshop, we will discuss these specialized problems and potential solutions, as well as offering suggestions based on our experiences in HPC cluster management. The topics will be based on the attendees' interest but may include OS deployment, software deployment, management tools, schedulers and resource managers, and customer issues.
Attendees should be admins of a compute cluster or interested in adminning this type of cluster. They should come prepared to discuss openly, in a round table setting, their admin experiences with this class of machine and the pros and cons of their existing cluster management tools.
author = {Clay England},
title = {Workshop 6: {HPC} Compute Cluster Workshop},
year = {2014},
address = {Seattle, WA},
publisher = {USENIX Association},
month = nov
}
