Tonglin Li

Biographical Sketch

Tonglin Li is a computer system engineer at NERSC. Before joining NERSC, he was a postdoctoral researcher at Oak Ridge National Laboratory. He received his Ph.D. degree in Computer Science from Illinois Institute of Technology in 2015. His research interests include distributed systems, large-scale storage systems, cloud computing, high-performance computing, and big data systems. 

Projects

ECP ExaHDF5: Independent HDF5 metadata update

In this ECP project, I'm working on building a serverless mechanism to maintain a consistent HDF5 namespace across participant processes. This will remove the limitation that applications must use collective operations to modify the namespace within an HDF5 file. 

EOD-HDF5: Advanced metadata and provenance management

I work on exploring strategies for capturing application metadata automatically when data is written to containers and for aggregating that metadata for collections of containers. I also conduct research on expanding the HDF5 interface to provide users with simpler ways to enter extensive metadata and provenance information through multi-attribute interfaces and configuration files.

Software release (approved by LBL Intellectual Property Office)

• Independent Metadata Update VOL Connector for HDF5 (IMM VOL).  
• https://bitbucket.hdfgroup.org/projects/HDF5VOL/repos/imm_vol/browse
• Provenance VOL Connector for HDF5 (H5Prov) 
• https://bitbucket.hdfgroup.org/projects/HDF5VOL/repos/provenance/browse
• Rootless Operations for MPI (RLO) 

• https://bitbucket.org/tonylbl/rootless-operations/src/master/

Publications and Presentations

• Independent metadata updating for large scale parallel I/O systems, Tonglin Li, Quincey Koziol, Houjun Tang, Jialin Liu, Suren Byna. Presented at NERSC Data Seminar. 
• H5Prov: I/O Performance Analysis of Science Applications Using HDF5 File-level Provenance, Tonglin Li, Quincey Koziol, Houjun Tang, Jialin Liu, Suren Byna, Cray User Group Conference (CUG), 2019
• Anycast: Rootless Broadcasting with MPI, Tonglin Li, Quincey Koziol, Houjun Tang, Suren Byna, Workshop on Exascale MPI (ExaMPI), 2018