University of Georgia deploys software to simplify and reduce cost of cluster management to aid research into carbohydrate-related diseases
Scali software has been deployed by the University of Georgia to cut system administration time and boost the reliability and robustness of its new Linux-based high performance computing (HPC) clusters.
The two commodity clusters, based on hardware selected from the HP unified cluster portfolio, have been installed in the Complex Carbohydrate Research Center (CCRC) at the university to help develop vaccines for many bacterial infections and diseases by simulating the reactions between carbohydrates and antibodies.
A combination of Scali Manage and Scali MPI Connect was selected for its capability to improve the functionality, reliability and administration involved in running critical cluster environments with high user demand.
In the past, the research laboratory invested considerable time and resource in building and maintaining its own clusters.
The fully supported and robust capabilities of Scali software solutions have delivered significant benefits to the CCRC, particularly quick installation, ease of management and improved workload monitoring, says Scali.
The CCRC scientists have been so impressed by the Scali software that they plan to use it across other clusters in the laboratory.
"Our experiments require the processing of massive amounts of data to deal with complicated computational codes.
"We needed cluster management software that would automate the time-consuming, but critical administration tasks.
"Our existing clustering management processes were draining our resources - what used to take us a week we can complete in a day using Scali software", said Robert Woods, CCRC, University of Georgia. "The software ensures the cluster is stable and runs smoothly so it has greatly improved our productivity. "It's freed up valuable time that we can now spend on experiments and furthering our research into vaccine design".
Woods's group at the CCRC is working to better understand the interactions between carbohydrates and proteins, to help it predict the energy of these interactions.
By simulating these interactions across the clusters, Woods and his team will determine the atomic level basis for carbohydrate recognition by the immune system and establish how to improve vaccines against bacterial and fungal infections.
"Our customers are telling us that in a typical cluster environment, the greatest cost is often attributed to the implementation and administration of ad-hoc software applications. "Scali's out-of-the-box software has enabled the CCRC to increase the efficiency of running its cluster solution while removing the headaches and administrative workload it has experienced in the past" said Bjorn Skare, president and CEO of Scali.
"This deployment is another example that highlights the need for powerful solutions that remove the complexity of cluster management".
The first deployment is an HP Linux-based eight-node cluster with dual Intel Itanium 2 processors, Myrinet interconnects, running Scali Manage and Scali MPI Connect, built to execute extremely demanding quantum calculations required for molecular simulations.
The second deployment is an HP Linux-based 64-node cluster with dual Intel Xeon processors, gigabit ethernet and Myrinet interconnects, running Scali Manage and Scali MPI Connect, to allow large molecular dynamic simulations to run in parallel.
Scali Manage enables companies to leverage a single solution for system installation, configuration, management and monitoring of dispersed clusters across the organisation.
Scali MPI Connect enables customers to take advantage of the leading interconnect hardware through a single MPI implementation.
