In this podcast, Rich Brueckner reports back from the LUG 2015 Lustre User Group Meeting. With something like 188 attendees this year, LUG reflects a user community that has come together to foster the world’s fastest parallel file system.
“Andreas Dilger from Intel presents an overview on the features currently under development for the upcoming Lustre 2.8 and 2.9 releases. This includes Layout Enhancement, Progressive File Layouts, Data-on-MDT, and improved single-client metadata and IO performance. In addition, several Lustre-specific ZFS improvements are also under development that will be available in this timeframe.”
“With the current Lustre Performance Monitoring Tool (LMT) no longer in active development, and the current version incompatible with DNE based Lustre 2.5 deployments, there is a critical need for a new set of tools delivering the same basic Lustre performance metrics with the added ability to work compatibly with contemporary releases of Lustre.”
“The Cray-Seagate partnership is helping expand the boundaries of what’s possible in large-scale, data-intensive computing, far beyond what we could have imagined just 10 years ago. This continued innovation using the Lustre open file system is helping assist data-intense applications critical to advancements in important industries around the world.”
“The combination of the ephemeral nature of the cloud and directly addressable archives such as S3 suggest novel methods for using the Lustre HSM interface. Persistent data sets in the cloud need to be managed independently from an ephemeral filesystem and compute resources. Managing datasets in the cloud could, for example, involves importing data from Amazon’s S3 back into a freshly-created Lustre filesystem, performing I/O intensive computations, and then persisting the datasets back to S3 before terminating the filesystem and compute resources. Alternatives for archive formats will also be discussed. AWS S3 will be used for concrete examples, but the general methods should be applicable to other cloud environments as well.”
“Large scale HPC IO is usually done either with a file per process or to a single shared file. Single shared file IO does not scale well in Lustre compared to file per process. This presentation from Cray’s Patrick Farrell will give details, examine the reasons for this, and explore existing and potential solutions. Group locks and a new feature, lock ahead, will be discussed in the context of strided IO.”
“Pleiades, one of the world’s most powerful supercomputers, represents NASA’s state-of-the-art technology for meeting the agency’s supercomputing requirements, enabling NASA scientists and engineers to conduct modeling and simulation for NASA missions. Powered by Lustre, this distributed-memory SGI ICE cluster is connected with InfiniBand in a dual-plane hypercube technology.”