In the second of a four-part series, Jay Etchings, director of operations for research computing and senior HPC architect at Arizona State University, lays out the concept of the campus cloud, a Research as a Service (RaaS) model with elements of Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and bare metal IaaS.
Embracing Hybrid Cloud/Conformant Cloud Concepts
“New technology is common, new thinking is rare.” – Sir Peter Blake
Proliferation of the campus cloud is nothing new or unique, but like any new pervasive technology it’s not the technology, it is what you do with it. Resource aggregation and rapid elastic scale is only practically possible in the hybrid cloud. Cloud-orientated architectures substantially increase the lifespan of a datacenter asset, but more importantly permit researchers to focus on research and discovery rather than wrestling the problems of scaling platforms.
The cloud-orientated architecture (COA) should not be confused with service-oriented architecture (SOA) concepts and cloud provider services as the latter typically only covers distributed computing items such as middleware. The University Hybrid cloud model, COA represents an amalgamation of clouds, living both in the private and public space. On-premise, used for pre-processing of massive amounts of raw source data and off-premise public clouds, both private and public where metadata and collaboratory data can live. Whereas little standards exist yet for cloud interoperability this model best suits university research computing.
The NGCC envisions a type of cloud service aimed at giving researchers and developers everything they need to develop applications, access Research as a Service (RaaS) whose subcomponents are the more traditional Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and bare metal IaaS. RaaS remove the administrative overhead of dealing with lower technology stack layers and reducing development overhead. Collections of software tools and libraries are installed, pre-configured, integrated and bundled into self-service options.
Within the cloud management platform there is great flexibility in toolsets for creation of a cloud service allowing true XaaS (XaaS is the hip collective term representing a number of things including X as a service, anything as a service or everything as a service). The COA model, synonymous with hybrid cloud in this writer’s minds unites PaaS, sharing of infrastructure between many researchers through secure multi-tenant delivery including VPC (Virtual Private Cloud) offering a lightweight alternative to a private cloud for remote collaboration where hybrid cloud resources exist within the secure network.
This model adopted by the NGCC and greater research computing initiative opens the door to on-demand resources, consumption metering and self-service to the performance computing world with functional automation of the whole technology stack.
The graphic below shows the logical architecture of the COA stack and RaaS components.
Leveraging the Internet2 Innovation Platform
“How do you innovate in a learning organization that is steeped with tradition? You do it by bringing in wave after wave after wave of innovation.”
– Paraphrase from Michael Crow, president of Arizona State University
What exactly is the Internet2 Innovation Platform and how does it transform research computing?
The Internet2 Innovation Platform is an orchestration of new technologies and services that provide a leading-edge, end-2-end architecture and a unique set of unified capabilities at the national, regional and campus level to create an environment for innovation in research and education. The Innovation Platform architecture creates an end-2-end networking environment that will enable new and unique applications in addition to supporting current science requirements.
Included in the Innovation Platform are key components such as bandwidth – massive amounts of bandwidth through a 100GE Layer 2 connections. This bandwidth is additionally enabled via software-defined networking (SDN) capabilities to support the development and deployment of new applications including a host of network function virtualization opportunities that create a pluggable, programmable research networks.
The traditional bottlenecks challenging hybrid campus cloud models for “X as a Service” are met with aggregation points to pass high bandwidth traffic providing performance monitoring / verification through implementation and support of a Science DMZ, a model developed by the Department of Energy’s ESnet. Participants commit to implementation of the three key components of the innovation platform architecture: 100GE Layer 2 connection, SDN and the Science DMZ model. The NGCC team in collaboration with advanced networking and systems experts have targeted the development of applications and application programming interfaces (API) for end-2-end optimization of research collaboration.
In brief the over-arching initiative includes network management modules facilitating point to point dynamic communications with application path optimization, integration and interface to Internet2’ Open Exchange Software Suite (OESS), collaboration with international research partners such as the ICTBioMed network, participation in larger collaborative commons models for conformant clouds with National Institute of Health (NIH), National Cancer Institute (NCI) Cancer Cloud Genomics pilots, NIH Big Data to Knowledge (BD2K) initiatives as well as full spectrum of programs such as Genomics Data Commons and the NIAID/NHGRI Human Microbiome Cloud. On the ASU immediate development path are security enhancements identifying and addressing essential challenges to building a robust firewall within the SDN/NFV layer that is stateful, distributed and portable (virtual). In addition to protocol enhancements robust security measures will be applied to SDN controllers creating secure enforcement kernels.
The diagram below outlines logical connections that are part of the Innovation Platform. As with any Innovation Platform a diverse range of connections and applications are expected but at minimum 100GE Advanced Layer 2 ports, SDN and a Science DMZ connecting campuses to be consistent characteristics of each Innovation Platform connection.
Part one of this series is available here. Stay tuned for parts three and four in the weeks ahead.
Director of Operations, Research Computing, and Senior HPC Architect at Arizona State University, Jay Etchings is a well-known industry professional with 20 years of progressively versatile, cross-platform experience in management of open systems architecture. With the bulk of a 10 year technical consulting career spent in gaming and connected lotteries, data relationship analysis has been a longtime passion for Etchings. He is well versed in all phases of cutting edge analytics and research computing. A former recovery audit contractor for the centers for Medicaid/ Medicare (CMS-RAC) positions him in alignment with the new ‘precision medicine’ healthcare field that is currently emerging.
Additional contribution provided by…
Dr. Kenneth Buetow also contributed to this article series. Buetow serves as director of Computational Sciences and Informatics program for Complex Adaptive Systems at Arizona State University (CAS@ASU) and is a professor in the School of Life Sciences in ASU’s College of Liberal Arts and Sciences. CAS@ASU is creating a Next Generation Cyber Capability (NGCC) to address the challenges and opportunities afforded by “Big Data” and the emergence of 4th Paradigm Data Science. This capability brings state-of-the-art computational approaches to CAS@ASU’s trans-disciplinary, use-inspired research efforts.