Hound: A Parallel Image Distribution System for Cluster Based on Docker

Current applications, consisting of multiple replicas, are packaged into lightweight containers with their execution dependencies. Considering the dominant impact of distribu-tion efficiency of gigantic images on container startup(e.g., dis-tributed deep learning application), the image "warm-u...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Journal of systems engineering and electronics 2023-08, Vol.34 (4), p.955-965
Hauptverfasser: Liu, Zijie, Li, Junjiang, Chen, Can, Zhang, Dengyin
Format: Artikel
Sprache:eng
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Current applications, consisting of multiple replicas, are packaged into lightweight containers with their execution dependencies. Considering the dominant impact of distribu-tion efficiency of gigantic images on container startup(e.g., dis-tributed deep learning application), the image "warm-up" tech-nique which prefetches images of these replicas to destina-tion nodes in the cluster is proposed. However, the current image "warm-up" technique solely focuses on identical image distribution, which fails to take effect when distributing different images to destination nodes. To address this problem, this paper proposes Hound, a simple but efficient cluster image dis-tribution system based on Docker. To support diverse image dis-tribution requests of cluster nodes, Hound additionally adopts node-level parallelism(i.e., downloading images to destination nodes in parallel) to further improve the efficiency of image distri-bution. The experimental results demonstrate Hound outper-forms Docker, kubernetes container runtime interface(CRI-O), and Docker-compose in terms of image distribution perfor-mance when cluster nodes request different images. Moreover, the high scalability of Hound is evaluated in the scenario of ten nodes.
ISSN:1004-4132
1004-4132
DOI:10.23919/JSEE.2023.000105