Businesses, research institutions and government agencies alike increasingly move to the hybrid cloud for the capability, flexibility and extensibility it offers. By maintaining a private cloud and augmenting it with a public cloud solution, organizations benefit from the best of both worlds. A private cloud keeps critical data and workloads aboard internally managed data centers, while the public cloud offers a cost-effective, elastic environment that offers scalability on-the-fly, security features and a marketplace of tools and services that provide additional features, including disaster recovery and more.
The health sciences industry recognizes many potential benefits the hybrid cloud offers. It provides medical researchers in both the government and private sector access to specialized on-premise infrastructure cloud platforms to tackle increasingly complex scientific endeavors like cancer research and genomic analysis. The ubiquitous nature of cloud platforms also ensures that medical professionals can access that patient data and applications anytime from anywhere. For these reasons and many more, the cloud offers a compelling solution indeed.
Despite these practical and functional upsides, health science organizations face unique challenges that hinder cloud adoption. In the United States, the Health Insurance Portability and Accountability Act (HIPAA) requires end-to-end security for any electronically transferred patient information. The European Union (EU) also enforces compliance laws to protect individuals’ privacy through the General Data Protection Regulation (GDPR). The GDPR also outlines rules that limit personal data transmission outside the EU. The healthcare industry must, therefore, accommodate all these requirements before the hybrid cloud offers a viable technology solution.
In November 2008, the San Diego Supercomputing Center (SDSC) Healthcare Cyberinfrastructure (CI) team set out to overcome these challenges — and more — with the innovative Sherlock Cloud. The Sherlock Cloud began with humble roots as an infrastructure as a service (IaaS) implementation.
Sandeep Chandra, division director for Healthcare CI at SDSC describes its origins. “Years ago, we received a government grant to create a secure system to move Medicaid data from an old mainframe to a new, more nimble enterprise solution, while ensuring adherence to federal security regulations. Because the on-premise system also incorporated Medicaid fraud investigation capability, we named the system Sherlock. We procured, configured and managed a substantial onsite compute and storage footprint to enable Sherlock Cloud’s functionality.”
Sherlock Cloud’s fledgling infrastructure grew into the largest solution within the University of California system, which met the Federal Information System Management Act (FISMA). Over time, Sherlock augmented their platform further to accommodate hundreds of National Institute of Standards and Technology (NIST) controls that dictate special compliance requirements, management processes, information control and system access protocols. When it came online in 2008, Sherlock Cloud provided an IaaS capability that complied with the federal Cloud First Policy intended to accelerate secure cloud computing adoption. These drivers ultimately resulted in SDSC’s innovative and specialized platform, which today offers a comprehensive hybrid cloud capability.
“Back then, not many organizations offered that type of end-to-end, secure, managed capability for healthcare data management,” noted Chandra. “After Sherlock’s successful implementation, other institutions approached us because they needed a similar end-to-end capability for scale, compliance, application hosting and data management. We recognized the community need and wanted to help them too.”
By architecting a more turnkey hybrid cloud platform that fully addresses compliance and security issues and provides value-added services on top, SDSC’s Healthcare CI efforts ultimately offered outside institutions a tool to rapidly and confidently reap the benefits of modern compute infrastructure to transform the way they conduct business or research. Hybrid cloud’s combined performance, reliability and dynamic scalability accommodated unique use cases and hefty workloads involving big data and analytics. As a result, hybrid cloud infrastructure offers government agencies, healthcare professionals and researchers a compute solution that facilitates breakthrough science and medicine.
But new endeavors often face challenges.
Chandra describes Sherlock Cloud’s growing pains. “The on-premise platform understandably required SDSC’s ongoing investment in staff expertise and physical infrastructure. We endeavored to recycle our storage and server hardware wherever possible, but also update it with the latest technologies like Intel Xeon Scalable processors, which offered added performance, reliability and scale to keep pace with today’s computing needs.”
In 2014, the Sherlock continued its sprawling growth to include a public cloud component from Amazon Web Services (AWS). AWS extended SDSC’s private cloud and offered increased ability to accommodate SDSC’s needs dynamically. In addition to storage, compute and networking solutions AWS offered SDSC out-of-the-box compliance, allowing SDSC to build innovative solutions upon the AWS services. The public cloud platform also offers SDSC supplemental new tools and libraries to speed deployment of applications and services. Plus, since the AWS and private cloud instances share a common Intel Xeon Scalable processor platform, the process of moving workloads between private and public cloud environments is simplified.
Chandra noted several immediate benefits of the new hybrid cloud solution. “By tapping public cloud vendors who run their systems on similarly robust hardware, we gain the best underlying technology — while simultaneously reducing our need for day-to-day systems management of SDSC’s on-site infrastructure,” he said. “Instead, we can focus on our core mission of building new and intelligent architectures and solutions to empower the industry.”
Today, SDSC has a number of clients benefitting from the ever-improving Sherlock platform, including academic organizations, researchers, medical institutions and industry partners. “They need IT experts to make cloud services less complex and more consumable, and we provide the means for them to do it,” Chandra said. “They choose our platform because it offers a simplified way to solve their scalability, security and compliance needs without the worry. While we work with each client to customize a solution for their unique needs, the platform gives them a substantial jump start for the process.”
Moving ahead, Chandra and his team foresee additional public cloud expansion for Sherlock. In addition to AWS, the SDSC team plans integration with Microsoft Azure and Google Cloud. Doing so offers their clients additional flexibility for unique use cases and applications.
For organizations considering the leap to a hybrid cloud solution, Chandra offers important takeaways from his team’s experience with Sherlock Cloud. “Sherlock’s on-premise, private cloud still makes sense in use cases involving a customized system configuration for a client or workloads involving heavy data analysis that depends on specialized hardware. However, the public cloud offers a cost-effective solution for many generalized, turnkey capabilities our clients need. We will continue relying on our established onsite systems. However, the benefits of a hybrid cloud solution make the choice a ‘no-brainer.’ For any organization facing challenges like ours, we strongly encourage them to consider a hybrid cloud environment. While it may take time to get all workloads tested and implemented fully through a hybrid cloud, the longer-term growth path offers a multitude of benefits.”
Send this to a friend