Presentations about Pawsey projects and best practices
TUESDAY 15 NOVEMBER
Case studies of ocean-related research enabled by supercomputers by Daniel Grimwood
An overview of three projects undertaken at Pawsey: The search for Malaysia Airlines MH370, modelling vessel drift for maritime safety, and building a 3D model of the HMAS Sydney and HSK Kormoran wrecks.
Australia’s National eResearch Collaboration Tools And Resources (NECTAR) and Pawsey by Jeremy Phillips
An introduction to the National eResearch Collaboration Tools And Resources (NeCTAR) – a collaborative Australian-based project, utilising OpenStack to provide cloud services to researchers. The presentation will also cover Pawsey’s contribution to NeCTAR.
WEDNESDAY 16 NOVEMBER
Tips and Tricks building an HPC Centre by Scott Atkinson
Lessons learnt during the design and construction of the Pawsey Supercomputing Centre. Tips and tricks that can be applied to both new and upgrade installations of HPC equipment. Includes common problems and their outcomes as well as what can be done to minimise the number of bugs in your installation. This “from the coal face” account is a must for anyone planning building works, upgrades or modifications.
Scanning the skies of Southern hemisphere by Mohsin Shaikh
The Pawsey Supercomputing Centre has an ongoing commitment to supporting the Radio astronomy and the Square Kilometer Array (SKA). The talk gives a brief overview of the SKA-1 and its precursors in Western Australia and related Supercomputing activity at the Centre.
THURSDAY 17 NOVEMBER
Experience with Large Spectrum Scale File Systems Backup & Restore at Pawsey by Jeffrey Jiang
Pawsey RDS File system is being accessed by researchers across Australia. It presently holds data from many scientific projects and disciplines. The file system usage has grown to nearly three Petabytes over and it is currently the largest Spectrum Scale file system used for research purposes in Australia. A full backup of this file system to a tape library via Spectrum Protect is performed periodically. During the presentation some performance previously identified pain points and the resolutions & techniques Pawsey has used to address them will be demonstrated, ranging from bottlenecks identification & turning to design change. The presentation will also aim to examine weaknesses of current solution and provide Pawsey plan for the future.
Mediaflux and Application Portal by Gene Schneider
The Pawsey Data Team use Mediaflux to provide an abstraction layer above the more complex underlying data storage mechanism, and provide a simplified user interface to reduce the user support workload on the team. In addition, an Application Portal has been built to handle Data Storage Requests from researchers. This Tim Tam talk will provide an overview of the tools used by the Pawsey Data Team.
Maali workshop by Chris Bording
One of the main operational challenges of HPC centres is maintaining numerous scientific applications in order to support a large and diverse user community. At the Pawsey Supercomputing Centre we have developed Maali, a lightweight automated system for managing a diverse set of optimized scientific libraries and applications on our HPC resources. Join Chris Bording for a demonstration.