3 Tips on How to Ease the Pain Points Surrounding Cloud Computing for Research

Subscribe for more like this



By Ananya Ravipati, Internet2 Research Cloud Engineer, and Amanda Tan, Internet2 Research Engagement Program Manager

Estimated reading time: 5 minutes

Ananya Ravipati (left) and Amanda Tan

Ananya Ravipati and Amanda Tan are two of several subject matter experts and program ambassadors who facilitate CLASS: Cloud Learning and Skills Sessions, an Internet2 program developed to fill the gaps in training for cloud and research computing. Ananya, a research cloud engineer at Internet2, is an in-house technical expert on Amazon Web Services, Google Cloud Platform, and Microsoft Azure. Amanda, a program manager for Internet2’s Research Engagement, is first and foremost a researcher who also helps bridge the communication gap between researchers and technical professionals. In this blog post, Ananya and Amanda offer three tips on how to ease common pain points surrounding cloud computing for research and invite you to sign up for the Fall 2021 CLASS Advanced Cohort. Early applications are due August 24. 

In and out of CLASS, we hear a lot from the research and education community about the pain points and hard lessons learned on the journey to adopting best practices for supporting research in the cloud. 

From effective cost planning to identity and access management, and steep initial learning curves to the sustainability considerations for the lifecycle of a given research project and beyond—the challenges of facilitating cloud computing for research are substantial. However,  they are not insurmountable.

Filling the [Vendor Agnostic] Training Gap

CLASS logo

CLASS teaches best practices on how to work in the cloud with a specific focus on research computing as opposed to enterprise IT. What’s particularly unique about the curriculum is that it’s vendor-agnostic. CLASS offers insight, exposure, and hands-on experiences into all three of the major cloud platforms: Amazon Web Services, Google Cloud Platform, and Microsoft Azure. 

The program starts off with structured training covering core concepts on the three platforms, including everything from cost management to data storage to containerization. Participants then explore solutions to real-world use cases that are applicable to their day-to-day job responsibilities. They also engage with a community of peers who face similar situations in similar environments. By discussing shared problems and crowd-sourcing solutions, they make collective progress on some of the recurring issues that otherwise take a lot more time to solve individually.

3 Tips on How to Ease Common Pain Points

Through communities of practice like CLASS, we gain a lot of insight into what works and what doesn’t. To share the wealth of that collective knowledge, here are three tips we can pass on to  other RCD professionals on how to ease some common pain points surrounding cloud computing for research:

  1. Remove barriers to entry related to cost and identity and access management by facilitating these in a structured, standardized way.

Hidden costs and steep learning curves for onboarding researchers to the cloud are big barriers to entry. Consult resources and opt into upskilling opportunities to deepen your understanding of best practices. Before you onboard researchers to their cloud accounts, walk them through the essentials like identity and cost management. Engage with them to understand why they need and want to use the cloud. This not only makes them more comfortable with the platform, but also makes the platform safer to use for all.

  1. Prototype first, and then build solutions for the long term.

When you recommend a cloud solution for a particular use case, build a proof-of-concept. This helps you get an estimate of the projected costs and potential risks. It also offers a head start on figuring out the support a research team might need to adapt to the platform. With that insight gained through the prototyping phase, you can focus on the long term—including understanding data egress costs and building a resilient and extensible cloud infrastructure. 

  1. Welcome and help facilitate the mindset change. Don’t let the inertia of business-as-usual dictate your organization’s cloud strategy. 

When you communicate with researchers about cloud computing, there needs to be a shift in mindset pertaining to the use of these platforms compared to any on-premises solution. Take, for example, ​​the concept of “scaling up.” In on-premises environments, that means buying new compute and storage hardware. Whereas, in the cloud, it means a few keystrokes to configure an increase in the number of instances in use or the size of the virtual machine. Establishing a shared understanding can play a key role in smooth and productive facilitation. It can also help you gain more insight into exactly what the researcher needs and avoid overprovisioning. 

BONUS TIP: Take advantage of the solutions that are already available, and don’t reinvent the wheel. When supporting a new research project in the cloud, always ask yourself: Are there use cases and workflows that you can reuse and adapt?

Looking for more tips, best practices, and insight from a community of peers about how to better leverage the cloud for research computing? Sign up for the CLASS Advanced Cohort in Fall 2021! Early applications, which receive 10% or more in tuition savings, are due August 24.