WORKSHOP: March 2020

The workshop is back again at the Francis Crick Institute in central London on the 3rd March. This year we welcome the newly appointed Director of e-Infrastructure for UKRI, James Hetherington. The programme includes speakers from a range of academic centres and representation from industry as well as regulars to this event such as CERN and the UK MetOffice.

Registration

We still have some places left if you’d like to register:

http://bit.ly/register-ukri-cloud-workshop2020

Programme

TimingPlenary / Strand A Sessions
Auditorium 2
Strand B Sessions
Auditorium 1
09:15Registration + teas/coffee
10:00Introduction – Philip Kershaw, STFC
Data & Computers & Code & People – plans for a distributed national digital research infrastructure – James Hetherington, Director of e-Infrastructure for UKRI
10:20Session 1a – Cloud Pilots
Chair: Philip Kershaw, STFC
Session 1b – Hybrid Cloud
Chair: Chris Woods, University of Bristol
The Sky is the limit: our journey to the Cloud – Stefano Angioni, University of BathDescribing portable applications with TOSCA -James DesLauriers, University of Westminster
Platform thrash! Getting Broad Hail / Spark software to work at scale in Sanger and Google and Sanger – Vivek Iyer, Wellcome Sanger InstitutePrototyping a hybrid cloud/on-premise CI/CD system for Nektar++, Christopher Cave-Ayland, Imperial College
Porting and performance of DiRAC HPC benchmarks on Oracle bare metal cloud – Andy Turner, EPCC/DiRACMillion core-hour rendering simulation on Cloud – Mitigating hybrid, scale, security and Cloud capacity challenges – Geoff Newell, AppsBroker
Question time
11:20Break – teas/coffee
11:45Session 2a – Organisation challenges in Cloud Computing – procurement, data governance, legal controls, cost controls

Chair: David Fergusson, University of Edinburgh
Session 2b – HPC in the cloud, performance and benchmarking

Chair: Steve Hindmarsh, Francis Crick Institute
Productive research on sensitive data using cloud-based secure research environments – Martin O’Reilly, Turing InstituteEarly results of AWS testing for research computing at the Crick – Steve Hindmarsh, Francis Crick Institute
Best Practice in Secure Research Computing – Stig Telfer, StackHPCTowards HPC resiliency, breakout and temporary transition in the Cloud – Cliff Addison, University of Liverpool
ASCLEPIOS – Health Care Security – Dr Gabriele Pierantoni, University of WestminsterCluster in the Cloud: Easy, Scalable and Heterogeneous – Matt Williams, University of Bristol
Question timeQuestion time
12:50Lunch – buffet lunch
13:50Session 3a – Development, deployment and operations for private and/or community clouds

Chair: Tim Cutts, Wellcome Sanger Institute
Session 3b – Lightning Talks


Chair: Simon Thompson, University of Birmingham
CERN Cloud Infrastructure Updates – Theodoros Tsioutsias, CERNMigrating a UK Genomics Organisation from public cloud to on-prem HPC Cloud – David Power, vScaler
Challenges in Scaling STFC Cloud – Alex Dibbo and Martin Summers, STFCTales From the Front Line Installing HPC in the Cloud – Owen Thomas, Red Oak Consulting
Autoscaling, Reservation, Contention and Preemption – the Coral Reef Cloud – Pierre Riteau,  John Garbutt and Steve Brasier, StackHPCARCHIVER – R&D for cloud-based Archiving & Long-Term Data Preservation Services – João Fernandes, CERN
JASMIN’s Cloudy path to Nirvana – Jonathan Churchill, STFCJupyterHub on Private Cloud – Vladimir Kiselev, Informatics Team Leader, Wellcome Sanger Institute
Convergence of AI and HPC Workloads – Implications for Cloud-Native and Hybrid Architectures – Gaurav Kaul, HPE
Question timeQuestion time
15:05Break – teas/coffee
15:35Session 4a – Addressing challenges integrating storage and data access interfaces for scientific workloads

Chair: Adam Huffman, Big Data Institute, University of Oxford
Session 4b – Demos




Chair: Andy Powell, Jisc
How to consume your data in the cloud without it costing the Earth? – CD Tiwari, EMBL-EBIA Universe from Nothing: Containerised OpenStack deployment using Kolla, Ansible and Kayobe – Stig Telfer, StackHPC
Building a distributed HPC-like environment using Federated Slurm and CephFS for processing astronomy data from the Euclid satellite – Mark Holliman, University of EdinburghCluster-as-a-Service for the JASMIN Cloud, Matt Pryor, STFC
Adopt a Cloud to change Research Computing at Diamond – Andrew Richards, Chris Reynolds, Diamond LightsourceInstanceHub: Simple provisioning of cloud instances for training – Paul Richmond, University of Sheffield
OpenFlightHPC – Stu Franks, Alces Flight Ltd
Question timeQuestion time
16:50Reconvene in Auditorium 2 …
17:00Final Plenary
Sum-up, feedback, next steps, cloud strategy for research community
17:10Reception – drinks, light refreshments
18:00Close