I manage the Systems group in Cloud and Information Services Lab at Microsoft. My group focuses on systems infrastructure aspects in the “Big data” space. Efforts range from rapid prototyping, to build production quality systems, releasing code to open source, and publishing papers in top Systems conferences. We work closely with Microsoft’s Big Data teams.
On a personal side, I am broadly interested in building storage and compute infrastructure for datacenter settings. I enjoy building and deploying systems in practice as well as releasing them as open source. In building these systems, my work leverages upon technology trends in datacenter computing.
My recent work has been in the area of resource management for Big data clusters. We have focused on a building a scale-out resource management substrate for big-data workloads. While the ideas are general, we have implemented our ideas on top of Apache Hadoop YARN:
Mercury (Hybrid Centralized/Distributed Scheduling; also, see YARN-2877)
Rayon (Rayon ships as part of Apache Hadoop 2.6; see YARN-1051)
Tetris (Packing tasks of Big data jobs to improve cluster efficiency)
Corral (Network-aware scheduling of Big data jobs)
Some of the previous systems I have built and released as open source projects are:
Kosmos distributed filesystem: I have designed/implemented/deployed (KFS) to manage PB’s of storage.
Sailfish: I have also designed/implemented Sailfish, a compute infrastructure which improves handling of intermediate data (i.e., “shuffle” phase in a Map-Reduce computation). Our results show that Sailfish can improve job completion times at scale by 20% to 5x.
At CISL, I am working on building Hadoop related services on Windows Azure. I also collaborate extensively with colleagues in MSR-Redmond.