Richie
Tan

Anvil REU Open OnDemand Dashboard Development for the Anvil Supercomputer Mathematical/Computation Sciences

Abstract profile. Full document pending author claim.

Authors:

Richie Tan

Date Created:

Not specified

Course Title:
Professor:

Not specified

About Paper:

Supercomputers like Anvil, the NSF-funded supercomputer built by Purdue University, provide powerful, cost- efficient computing resources for researchers nation-wide. This REU project focuses on bridging the gap between the Linux terminal and the web-based Open OnDemand user experiences on Anvil by combining responsive, informative, and user-friendly dashboard design with detailed job management and performance metrics, enabling productive, streamlined research for researchers with varying amounts of computing experience. The updated dashboard homepage developed in this project introduced widgets displaying users' service unit credit and disk usages, important Anvil announcements, and partition statuses and resource load to provide researchers a quick, centralized overview of key information. By querying Slurm, an open-source job scheduling and accounting software used on Anvil, the dashboard transforms difficult-to-parse data from Anvil into easy-to-understand tables and charts without the user ever interacting with the terminal and Linux commands. The improved dashboard concisely formats key information about a researcher's jobs while offering the option to view more detailed data, allowing researchers to quickly analyze their jobs without being overwhelmed by irrelevant information. Further information regarding each job's memory, CPU, GPU, and time utilizations will provide researchers more comprehensive insights to their jobs on Anvil. The new Anvil dashboard will lower the barrier of entry to high performance computing for researchers without sacrificing important functionality and metrics traditionally provided by Anvil's terminal interface. Keywords: HPC; Open OnDemand; Anvil; Ruby on Rails; Slurm

Source:

Purdue University / 2024

Topics:

No topics listed

Co-authors:

Richie Tan

0