Hound: Causal Learning for Datacenter-scale Straggler Diagnosis – Dr. Benjamin Lee, Duke University

Stragglers are exceptionally slow tasks within a job that delay its completion. Stragglers, which are uncommon within a single job, are pervasive in datacenters with many jobs. We present Hound, a framework that models task latency using system conditions to reveal causes of stragglers for each job. Hound discovers recurring causes across jobs by constructing […]

Continue Reading