Watch our success story
Dr. Borries Luberacki, Head of HPC Operations at the GMI, begins: “New approaches like next-generation sequencing enable scientists to dig into plant DNA quickly and more cheaply than traditional methods—resulting in a rapid growth of data.”
In addition, advances in medical and biological imaging technology mean that researchers at the GMI need to process more data than ever before.
Dr. Borries Luberacki comments: “Today, many researchers want quick access to HPC resources to analyze large, complex data sets. But with more scientists eager to analyze more data with new, modern computational tools, it was becoming increasingly difficult for our small team to prioritize and manage it all.”
Dr. Borries Luberacki notes: “Since our HPC cluster runs SUSE Linux Enterprise Server as the operating system, SUSE OpenStack Cloud was the natural choice to expand our computing capabilities. We wanted an enterprise solution from a vendor we could trust and, having partnered with SUSE for many years, we knew that we could rely on them for support.”
Working with the SUSE team, the GMI implemented SUSE OpenStack Cloud alongside its HPC cluster. Dr. Borries Luberacki says, “The SUSE team was very helpful throughout the implementation, particularly when we ran into hardware complexities. Whenever we have any questions, we can get help from SUSE right away.”
With SUSE OpenStack Cloud, the GMI established a software-defined data center and can quickly deploy virtual machines (VMs) in a private cloud close to the institute’s HPC cluster. The GMI uses the technology to provide fast, flexible Infrastructure-as-a-Service (IaaS) offerings directly to scientists.
Dr. Borries Luberacki remarks: “When scientists wanted to try out a new analytics application in the past, it would take us weeks, sometimes months, to set up the software stack on the HPC cluster. Because it took so long, we’d often find that by the time we’d finished setting up the solution, it was no longer needed.
“With SUSE OpenStack Cloud, we can provide scientists with their own VMs in our private cloud in as little as 10 minutes, enabling them to deploy the scientific workflows themselves, as and when they want. This self-service model means that scientists have greater control, flexibility and agility over how and when they use HPC to support their research.”
Thanks to the platform’s easy-to-navigate graphical interface, scientists can quickly and easily launch VMs to deploy their workloads without involving the HPC team
Dr. Borries Luberacki comments: “Now, when scientists want computing resources to analyze next-generation sequencing and biological imaging data sets with the latest computational tools, they can get started themselves without delay. Not only does this support scientific experimentation and innovation, it also greatly reduces the HPC team’s workload. With scientists able to deploy scientific workflows themselves, we only have to provide the basic infrastructure. They can test and evaluate tools independently, and only turn to us when they need additional support.”
He concludes: “We’re already planning to implement SUSE OpenStack Cloud 7 to take advantage of the latest Container-as-a-Service capabilities. Containers make sharing complete applications including all necessary dependencies much easier. Enabling our scientists to take full advantage of new computational tools will further support cutting-edge research.”