GX-Plug: a Middleware for Plugging Accelerators to Distributed Graph Processing

Recently, research communities highlight the neces-sity of formulating a scalability continuum for large-scale graph processing, which gains the scale-out benefits from distributed graph systems, and the scale-up benefits from high-performance accelerators. To this end, we propose a middleware, call...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Data engineering s. 2682 - 2694
Hlavní autoři: Zou, Kai, Xie, Xike, Li, Qi, Kong, Deyu
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.05.2022
Témata:
accelerators
Computational modeling
Data engineering
Distributed graph systems
Middleware
Pipelines
Programming
Runtime environment
Scalability
Synchronization
ISSN:2375-026X
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Recently, research communities highlight the neces-sity of formulating a scalability continuum for large-scale graph processing, which gains the scale-out benefits from distributed graph systems, and the scale-up benefits from high-performance accelerators. To this end, we propose a middleware, called the GX-plug, for the ease of integrating the merits of both. As a middleware, the GX-plug is versatile in supporting different runtime environments, computation models, and programming models. More, for improving the middleware performance, we study a series of techniques, including pipeline shuffle, synchro-nization caching and skipping, and workload balancing, for intra-, inter-, and beyond-iteration optimizations, respectively. Exper-iments show that our middleware efficiently plugs accelerators to representative distributed graph systems, e.g., GraphX and Powergraph, with up-to 20x acceleration ratio.
ISSN:2375-026X
DOI:10.1109/ICDE53745.2022.00246