A Runtime System for Interruptible Query Processing: When Incremental Computing Meets Fine-Grained Parallelism - Artifact
Year: 2024
Type: article
Abstract: The goal of PitStop is to use lazy evaluation to improve both throughput and operation latency for data processing systems. PitStop allows for queries and updates to be temporarily halted mid-traversal and continued at a later time. Though it may seem counter intuitive, this pausing allows for a fine-grained parallelism that helps to balance workload and alleviate workload fluctuation. This design also allows for further query optimization through the use of techniques such as batching and fusion.
Authors: Eymer, Jeff, Dexter, Phillip, Raskind, Joseph, Liu, Yu David
Institution: Binghamton University
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W6893037840
