Research and ACM SRC Posters Archive

Extending I/O Observability to Explainability for Targeted Optimization


Poster Type: Research Posters

Author: Neeraj Rajesh (Illinois Tech), Sarah Neuwirth (Johannes Gutenberg University Mainz), Jay Lofstead (Sandia National Laboratories), Anthony Kougkas (Illinois Tech), Xian-He Sun (Illinois Tech), Nikoli Dryden (Lawrence Livermore National Laboratory (LLNL)), Hariharan Devarajan (Lawrence Livermore National Laboratory (LLNL))

Abstract: HPC I/O profilers such as Darshan report POSIX-level metrics but do not trace the kernel stack: VFS, page cache, block layer, NVMe driver, and storage media. We present PEARL, an eBPF-based cross-layer tracing plugin for DataCrumbs, that builds one causal timeline per request from syscall to disk completion. We evaluate PEARL on Chameleon Cloud with IOR workloads. Page-cache faults account for 82% to 93% of kernel I/O time, a signal Darshan cannot capture. PEARL traces the same requests at 9.0 to 15.7 microseconds where Darshan reports 0.45 to 5.88 milliseconds. Overhead is negligible and within system noise. PEARL moves profiling from reporting what happened to explaining why, enabling targeted evidence-based optimization.

Best Poster Finalist (BP): no
Poster: PDF
Poster Summary: PDF


Back to Poster Archive Listing