Schedule subject to change
| Date/Time(EST) | Session | Session Dur. (minutes) | Location | Session Title | Description/Theme | Presenters |
|---|---|---|---|---|---|---|
| Monday 11/15 | ||||||
| 6:00-7:00 PM | Reception | 60 | Statler Hotel Terrace Lounge | |||
| Tuesday 11/16 | ||||||
| 8:00 AM | Registration/ Breakfast | 30 | Conference Foyer | |||
| 8:30 AM | Opening/Survey Results (plus bonus content) | 30 | Slides | Discussion of trends exposed by annual cloud survey of registrants | ||
| 9:00 AM | Presentation 1 | 30 | Slides | The Big Lie: Cloud Computing | In 2011, I started evangelizing for the Cloud. It’s been 10 years, so the future I promoted has come true, right? I’ll revisit the vision we imagined, what actually happened, and explore the question “Is cloud computing the biggest disappointment for higher education this century?” | Bill Wrobleski – Penn State |
| 9:30 AM | Presentation 2 | 30 | Slides | Lessons Learned from AWS Student Data Lake POC | As part of Lehigh’s Strategic Data Analytics Initiative, a small team from Library and Technology Services, Institutional Research, and Provost Office collaborated with Amazon to deploy to data lake proof of concept around student data. The goal of the project was to determine if AWS was the right platform for Lehigh to deploy a strategic data lake initially around student data but with capabilities to scale far beyond that. | James Monek & John Zekind – Lehigh University |
| 10:00 AM | Panel 1 | 45 | Amphitheater | Advances in VDI in Higher Ed | Representatives of institutions with a variety of approaches to virtual desktop and on-demand application delivery will share their successes and challenges. | |
| Noah Abrahamson – Stanford (moderator) | ||||||
| Kris Barth – Cornell University | ||||||
| Damian Doyle – UMBC | ||||||
| James Smith – University of Notre Dame | ||||||
| Jonathan White – UCF | ||||||
| 10:45 AM | break | 15 | Conference Foyer | |||
| 11:00 AM | Lightning Talks | 60 | Slides || Recording | |||
| Custom Biomedical Cloud Training from the NIH STRIDES Initiative | The National Institutes of Health (NIH) Science and Technology Research Infrastructure for Discovery, Experimentation, and Sustainability (STRIDES) initiative provides NIH, and NIH-funded, researchers access to cloud training opportunities from cloud providers. Adoption of cloud technologies and analytics specifically for biomedical research has created the demand for targeted trainings currently not broadly available. To address this need, we created two, one-day training courses introducing fundamental life science tools native to both Google Cloud and Amazon Web Services, with learning activities customized to computational biomedical researchers. Satisfaction scores from this course show that creating targeted cloud training, incorporating the use of biomedically relevant data sets, is impactful. | Philip Meacham, Ph.D. – NIH | ||||
| Cloud Cost Oversight for Projects and Programs under the STRIDES Initiative | Transparency (easy access to and intuitive reporting of the types of resources utilized for and costs incurred against a cloud billing account) and awareness (timely notification of incurred costs that exceeds anticipated norms) in the oversight of cloud billing accounts remains a vital requirement for investigators and others responsible for the financial management of individual research projects and large portfolios alike. To address this need, the National Institutes of Health (NIH) Science and Technology Research Infrastructure for Discovery, Experimentation, and Sustainability (STRIDES) Initiative has implemented a customized, multi-cloud cost and usage dashboard using Tableau to provide better insight into cost trends for data storage/utilization and computation supporting NIH-funded biomedical research. In this presentation, we will describe the requirements the STRIDES Team has received from our hundreds of participants and how those requirements drove implementation of the current cost reporting features we provide to NIH and NIH-funded research institutions, as well as a summary of additional features the STRIDES Team is considering for future development in the dashboard and cost reporting arena. |
Yanik Wynter – NIH | ||||
| NIH Approaches to Cybersecurity and Cloud | As NIH systems and applications are migrated to the cloud from on-premises or are natively developed in the cloud, NIH cybersecurity teams must also make this transition from on-premises security practices to securing the cloud and the resources they deploy in the cloud. To ease this transition, NIH’s (and NIH’s Institute’s and Center’s, or IC’s) security solutions can be leveraged to gain visibility to their cloud environments and to help secure their cloud service provider accounts and deployments. In this lightning talk, we plan to briefly discuss the cloud security capabilities of common enterprise solutions, such as Tenable/Nessus and Splunk, as well as how native cloud service provider (CSP) security solutions can augment these enterprise security capabilities. |
Carlo Espiritu – NIH | ||||
| Cloud Wrangling at NIH | Organizing and managing multiple cloud infrastructure and platform service providers in an institution-/enterprise-wide fashion is a daunting effort. Whether the concept of an organization was built in from the beginning or a bolt-on later, you may find your organization has more than one tenant within the same provider. Or procurements have been made using different contracts or agreements, via different resellers. Or there are collections of accounts from a cloud provider that have conflicting standards as your organization’s cloud maturity has evolved. This quick talk will cover these challenges and describe how those of us at the National Institutes of Health have begun to wrangle them into manageable structures. | Joel Peterson – NIH | ||||
| Cloud Golf | Minimize wall clock time and keystrokes (hence ‘golf’), maximize awesome functional research machinery. The purpose of the session is to take one step beyond “let me tell you what is possible” to “you can build this in 20 or 30 minutes yourself.” Emphasis on research, emphasis on it actually works. An example entry would be a procedural in a GitHub repository that (assuming one has a cloud account already established) contains a walkthrough for creating a serverless sudoku solver. This checks the boxes of [x] cloud implementation, [x] serverless API, [x] non-trivial (recursive) calculation as a service, and [x] does in under one second what a human might accomplish in hours. | Rob Fatland – CloudBank | ||||
| Trudging toward compliance: NIST 800-181 in Azure | This talk will describe how we’re using Azure policy, blueprints, technical and non-technical controls, and a cross-functional team to create a virtual desktop environment for our secure data enclave in Azure. | Shelley Rossell – Univ Chicago | ||||
| IT’S A PILOT! Legitimating experimental work in your cloud | Has this happened to you? You have a cloud that’s not ready to take payments, or not ready to be used by the masses, or you don’t know how to do things right — but something comes through that you shouldn’t say no to. UChicago has handled these situations with a lightweight pilot framework. Both the cloud people and the user agree in advance, in writing, on what both parties will get from the situation. It has helped to set expectations about the unexpected and ultimately get unstuck. | Cornelia Bailey – Univ Chicago | ||||
| CLASS: Cloud Learning and Skills Sessions | This session is a talk about the CLASS program by Internet2 that is aimed to train the research computing and data (RCD) professionals to use Cloud platforms effectively. | Ananya Ravipati – Internet2 | ||||
| Give the People What They Want (Assuming they know what it is) | Over the summer of 2021 I prompted a conversation with the higher ed cloud community about what vendors and services, in support of the primary cloud vendors, they would like to see community contracting collaboration effort around. In this lightning talk I’ll report on how that all worked out and what we learned from it. | Bob Flynn – Internet2 | ||||
| 12:00 PM | Lunch | 60 | Taylor/Rowe | |||
| 1:00 PM | Researcher 1 | 60 | Recording | A Hitchhiker’s Guide to Big Data Based Biomarkers | Few biomarkers derived from genome scale data have translated into improved clinical classification of cancer subtypes, in spite of the wealth of available genome-wide studies and of the corresponding application of numerous statistical algorithms. This widespread shortcoming also derives from the pervasive use of “off the shelf” algorithms and machine learning techniques developed for image classification and language processing, which are naïve of the underlying biology of the system. Systematically leveraging prior information about biological networks (and more) can simultaneously constrain the search for predictive models to those with a potentially mechanistic justification and overcome the technical limitations inherent in “tabula rasa” statistical learning. Ultimately, “hardwiring” mechanisms into the predictive models is a “win-win”: on the biological side it enhances the translational value of the classifiers by hypothesizing causal explanations for disease phenotypes; on the statistical side it forcefully addresses the “curse of dimensionality”, increasing robustness by limiting the model space. | Luigi Marchionni, Associate Professor of Pathology – Weill-Cornell Medicine |
| 2:00 PM | Workshop | 90 | Slides || Recording | Avoiding “Networking” in Your Cloud Adoption Strategy | Establishing secure, private networking between campus and the cloud has long been a barrier to cloud adoption in higher education. To help institutions adopt cloud IaaS and PaaS technologies effectively and avoid making your cloud environment dependent on your campus network, WUSTL’s John Bailey proposes an “Internet first” approach when designing new cloud workloads to meet specific business needs. By building cloud “islands” (John’s term for cloud IaaS and PaaS deployments that do NOT require private network connections back to campus) we can simplify the architecture of many workloads while at the same time making them more robust.In this workshop, John and colleagues, together with engineers from AWS, Google, and Microsoft, will lead the Forum through a discussion of the wisdom of this approach, where to apply it, and how to implement it on the three platforms. | |
| John Bailey – Washington University in St. Louis | ||||||
| James Monek – Lehigh University | ||||||
| Erik Lundberg – University of Washington | ||||||
| Kevin Murakoshi – AWS | ||||||
| Kristy Patullo – Google | ||||||
| Ken Hoover – Microsoft | ||||||
| 3:30 PM | Break | 15 | Conference Foyer | |||
| 3:45 PM | Workshop continues | 75 | Slides | Avoiding “Networking” in Your Cloud Adoption Strategy | ||
| 5:30-8:30 PM | Dinner | Statler Hotel Terrace Restaurant | ||||
| Date/Time(EST) | Session | Session Dur. (minutes) | Location | Session Title | Description/Theme | Presenters |
| Wednesday 11/17 | ||||||
| 8:00 AM | Breakfast | 30 | Conference Foyer | |||
| 8:30 AM | Presentation 3 | 30 | Slides | Automating Governance At Scale with Customizations for AWS Control Tower | AWS Control Tower provides built-in guardrails and governance features, but did you know it can also be leveraged to automate the deployment of custom resources and policies across your organization? Using the “Customizations for AWS Control Tower” solution from AWS, along with a little help from DLT and a consulting partner, Northwestern is able to automate the deployment of network connections, cost monitors, and other resources to new and existing AWS accounts. | Matthew Rich & Dan Landerman – Northwestern University |
| 9:00 AM | Presentation 4 | 30 | Slide | Same Research, Different… Everything? | Evolving research at the speed of Cloud! How do you continue to support medical research when the demands for tools, access, collaboration, customized environments, connection options and increased compliance converge on your Information Services team? At Penn Medicine, the University of Pennsylvania and its Health System are turning to cloud resources for solutions, but it’s not simple. Complexity is the order of the day as solutions are proposed, tested, prototyped, and presented to the world-class investigators we serve. | Rikki Godshall – Penn Medicine |
| 9:30 AM | Presentation 5 | 30 | Slide | The Case for Compliant Cloud Computing | Meeting modern compliance requirements for research can be a demanding task, but the agility and flexibility of cloud technologies give us ways to support researchers’ needs more efficiently than ever before. What goes into creating these compliant cloud architectures? How do we partner with campus stakeholders to build a support structure for researchers? And crucially, is the cloud value proposition truly there, or should we resign ourselves to racking and stacking in search of ROI? Notre Dame is tackling these questions and more as we develop AWS environments to support CUI workloads, HIPAA research, and anticipate future requirements for CMMC and beyond. Join us for a conversation on where we’ve been, where want to be, and how we can all work together to get there. | Brandon Rich & James Smith – Notre Dame |
| 10:00 AM | Panel 2 | 45 | Slides | United We Stand, Divided We Get What We Want: The Pros and Cons of Direct vs. Collective Agreements | Programs like Internet2 NET+ help us leverage the community’s desire to combine our buying power, common architecture, etc. in establishing agreements with the cloud vendors. This returns tangible benefits in cost and terms, but it sometimes constrains our ability to meet practical implementation requirements. What tradeoffs in terms of control are schools willing to make? Where is the right balance between the power of a collective agreement and the autonomy of a direct agreement? | Sean O’Brien – Internet2 (moderator) |
| Sarah Christen – Cornell University | ||||||
| Damian Doyle – UMBC | ||||||
| Rick Rhoades – PSU | ||||||
| 10:45 AM | break | 15 | Conference Foyer | |||
| 11:00 AM | Presentation 6 | 30 | Slides || Recording | Managing Cloud Storage – the New Reality | The era of unlimited storage in SaaS platforms is over. This places new requirements on us to actively manage storage, meaning we need to build retention policies, automated pipelines, tiering strategies, and more.Where will that data live in the future? Presumably, at least a good deal of it will end up migrating to public cloud storage platforms. Are the cloud teams at our institutions ready to take on the migration and active management of data at that scale? What do we need to do to prepare for these scenarios? | Oren Sreebny – Internet2, Helen Hockx-Yu – Notre Dame |
| 11:30 AM | Presentation 7 | 30 | Sldies || Demo || Recording | Simplifying Research in the Cloud | Advanced Research Computing at the University of British Columbia has been consistently exploring ways to bring the power of the Public Cloud to researchers of all backgrounds and technical skill levels. To this end, we will present our implementation and experiences of the RONIN platform. RONIN is an incredibly simplistic web application that allows researchers to launch complex compute resources leveraging AWS Public Cloud infrastructure within minutes.We will explore the use of RONIN to deploy common applications (such as R Studio and Jupyter Notebooks), HPC Clusters, Object storage, and remote desktops for visualization. Also explored will be how researchers can optimize their usage of Public Cloud resources using pre-emptible instances and machine scheduling. We will also discuss how implementing such a platform significantly reduced barriers and increased adoption of the Public Cloud by researchers. | Ken Bigelow & Venkat Mahadevan – University of British Columbia |
| 12:00 PM | Lunch | 60 | Taylor/Rowe | |||
| 1:00 PM | Researcher 2 | 60 | Recording | Lessons Learned from VUMC’s First End-to-End Cloud Project | In 2020 and 2021, the Vanderbilt University Medical Center (VUMC) migrated many core data resources in Health IT from on-premise to cloud. In this presentation, we will describe our NCI-sponsored research project that VUMC used as a pilot after the cloud migration called “Follow-up Interactive Long-Term Expert Ranking” (FILTER). The focus of discussion will be both on the project as well as the decision-making process that made the project successful. We will also share lessons learned from the pilot and how those are shaping the VUMC cloud policy going forward. | Dr. Travis Osterman – Vanderbilt University Medical Center |
| 2:00 PM | Presentation 8 | 30 | Slides | AWS Account Lifecycle Management | We will talk about the lifecycle of an AWS account at Iowa, including provisioning, inventory, and how they’re terminated. We will discuss code we’ve written to compliment some of the policy guardrails AWS provides natively and highlight our API-based inventory system for tracking our 100+ accounts. Finally, we’ll discuss where we’d like to go in the future. | Chris Lawrence – University of Iowa |
| 2:30 PM | Unconference Slot 1 | 45 | Amphitheater | Organize Unconference | Unconf – Organize | Noah Abrahamson – Stanford University (Instigator) |
| 3:15 PM | Break | 15 | Conference Foyer | |||
| 3:30 PM | Unconference Slot 2 | 45 | Amphitheater & Conference Foyer (as assigned) | Unconference Session 1 | Unconf – TBD during Organize Session | |
| 4:15 PM | Unconference Slot 3 | 45 | Amphitheater & Conference Foyer (as assigned) | Unconference Session 2 | Unconf – TBD during Organize Session | |
| 5:00 PM | Unconference Slot 4 | 45 | Amphitheater & Conference Foyer (as assigned) | Cloudbursts | Unconf – Report Out | |
| 5:45 PM | Dinner – on your own | |||||
| Date/Time(EST) | Session | Session Dur. (minutes) | Location | Session Title | Description/Theme | Presenters |
| Thursday 11/18 | ||||||
| 8:00 AM | Breakfast | 45 | Conference Foyer | |||
| 8:45 AM | Presentation 9 | 30 | Slides | The Challenges and Success of Supporting Researchers on the Cloud | In 2019, Emory University launched its AWS at Emory service with its primary audience of researchers and their teams. In addition to the AWS architecture, security controls, integration with financial systems, and multi-step automation of accounts, Emory dedicated a technical team to help investigators determine whether or not they are a good fit for the Cloud and, if so, help them transition to our AWS at Emory service. Over the two years, we have over a hundred accounts created with services spanning from compute to data lakes. We have had lots of bumps, walls, and wins. In this session, we share our challenges, our successes, and strategy that includes active engagements and an on-going Community of Practice. | Circe Tsui – Emory University |
| 9:15 AM | Presentation 10 | 30 | Slides | Exploring Clouds for Acceleration of Science | Exploring Clouds for Acceleration of Science, or E-CAS, is a project funded by the National Science Foundation (NSF) that is being managed by Internet2 in collaboration with commercial cloud providers. The project solicited proposals from researchers representing a variety of academic disciplines and was designed to better understand how cloud capabilities can help in advancement of various sciences. This session is about the scope of the final two projects, takeaways from facilitating this project, blockers they faced to get to using the platforms and the key achievements that are made possible by the scalability of cloud platforms. | Ananya Ravipati – Internet2 |
| 9:45 AM | Break | 15 | Conference Foyer | |||
| 10:00 AM | Researcher 3 | 60 | Recording | Implementing Genomic Medicine at the HGSC | Genomic medicine provides a way forward to improve healthcare. However, the obstacles are numerous, including difficulties in interpreting genetic data in the context of rapidly-changing biological knowledge, integrating genetic data into electronic health records, and ensuring that reported findings are up to date. In this talk, we describe a range of efforts taking place in the clinical laboratory at the Human Genome Sequencing Center to address these challenges in the context of multiple high-throughput clinical sequencing projects like the All of Us Research Program, HeartCare and the eMERGE Network. A key component of this effort has been taking advantage of cloud technologies both to carry out data processing at scale and to handle protected health information. We have applied these tools and pipelines for the generation and delivery of over 15,000 clinical genomic reports and the primary analysis of over 50,000 genomic samples across multiple projects. | Prof. Eric Venner, Associate Professor, Division Director for Bioinformatics – Human Genome Sequencing Center – Clinical Lab. Baylor College of Medicine |
| 11:00 AM | Presentation 11 | 30 | Slides || Recording | The Creation and Operations of the Cloud Financial Analyst Position at UC San Diego Cloud Services and CloudBank | As UC San Diego ITS moved more and more services to the cloud, cost were spiraling out of control. An Enterprise Architect was assigned to get a handle on spending and to set up a sustainable approach to managing cloud cost growth and optimization. At the same time, UC San Diego joined with UC Berkeley and the University of Washington to create CloudBank to serve NSF researcher use of the cloud. This led to the creation and recruitment of a Cloud Finance Analyst, combining tech and finance skills. We will discuss the work the position does to optimize and model sustainable workloads for UCSD and CloudBank. Some examples are using optimization software to create reports for individual accounts, and the follow up needed to execute them; and coordinating with architects, engineers, our reseller, and service providers to craft the right balance of on-demand, reserved instances, and savings plans. | Declan Fleming & Jessica Respicio – University of California San Diego |
| 11:30 AM | Presentation 12 | 30 | Slides || Recording | TeamWow! | You’ll be saying wow every time you have a team meeting. That’s because your team holds 33% more people than you actually have!” – Rick Rhoades, Satisfied Customer. This presentation will discuss the partnership between the Penn State IT Talent Management group and Penn State Cloud Services. When the pandemic hit, Cloud Services was unable to continue with hiring their open position. With a large backlog of project work, the team was facing significant delays in completing service development for their cloud providers: AWS, Azure and GCP. Thankfully, the IT Talent Management team was preparing to launch their pilot effort around a new concept called the “Talent Share Program.” The Talent Share Program is a blended learning opportunity that combines prerequisite training with an on-the-job application allowing employees to learn about an area of interest. This program has been key to our continued service development, internal talent pipeline, and a unique way to assist building cloud fluency across the organization. |
Rick Rhoades – Penn State |
| 12:00 PM | Break & Box Lunch | 15 | Taylor/Rowe | |||
| 12:15 PM | Presentation 13 | 30 | Slides | Developing a Cloud Secure Enclave for CUI | This session will give an overview of WashU’s initiative that has established a secure enclave in Azure to store and transmit CUI (Controlled Unclassified Information.) We will discuss reasons for a separate secure enclave (as opposed to securing everything), architectural considerations, vendor selection considerations, platforms/tools selected, and lessons learned. The session will cover both technical details as well as management considerations for such a project. | John Bailey – Washington University in St. Louis |
| 12:45 PM | Presentation 14 | 30 | Slides | DevOps Conference Wrap-up | On the afternoons of June 2-4, the first Cloud Forum-sponsored Higher-Ed DevOps Conference virtually welcomed ~120 attendees from 56 institutions. The conference focused on how to use new technologies and capabilities to deliver value more quickly and more reliably to our organizations. We had several talks about a variety of topics and then an entire afternoon set aside as an unconference to talk about whatever attendees wanted to learn.As conference organizers, we learned a lot about running an event in 2021. But, we also heard several common themes and concerns. In this session, we’re going to talk about these takeaways, next steps for the conference/DevOps focus, and a few potential areas for strategic focus, especially as we move to a world where we are competing against a global (or at least national) marketplace for employees. | Michael Irwin – Virginia Tech & Jay D. Hulslander – Cornell University |
| 1:15 PM | Closing | 30 | Amphitheater | Cloud Forum Wrap-up | Sarah Christen – Cornell, Bob Flynn – Internet2 | |
| 1:30 PM | Adjourn | Conference Foyer |