Scalable computing in Java with PCJ Library. Improved collective operations

This work presents an introduction to the Data Intensive Scalable Computing (DISC) approach. This paradigm represents a valuable effort to tackle the large amount of data produced by several ordinary applications. Therefore, subjects such as characterization of big data and storage approaches, in addition to brief comparison between HPC and DISC are differentiated highlight.

Download Full-text

On Task Assignment in Data Intensive Scalable Computing

Job Scheduling Strategies for Parallel Processing - Lecture Notes in Computer Science ◽

10.1007/978-3-662-43779-7_8 ◽

2014 ◽

pp. 136-155

Author(s):

Giovanni Agosta ◽

Gerardo Pelosi ◽

Ettore Speziale

Keyword(s):

Task Assignment ◽

Scalable Computing ◽

Data Intensive

Download Full-text

GalaxyCloudRunner: enhancing scalable computing for Galaxy

10.1101/2020.05.28.121772 ◽

2020 ◽

Author(s):

N Goonasekera ◽

A Mahmoud ◽

J Chilton ◽

E Afgan

Keyword(s):

Source Code ◽

Supplementary Information ◽

Scalable Computing ◽

Link Type ◽

Cloud Providers ◽

Galaxy Server ◽

Cloud Resources

AbstractSummaryThe existence of more than 100 public Galaxy servers with service quotas is indicative of the need for an increased availability of compute resources for Galaxy to use. The GalaxyCloudRunner enables a Galaxy server to easily expand its available compute capacity by sending user jobs to cloud resources. User jobs are routed to the acquired resources based on a set of configurable rules and the resources can be dynamically acquired from any of 4 popular cloud providers (AWS, Azure, GCP, or OpenStack) in an automated fashion.Availability and implementationGalaxyCloudRunner is implemented in Python and leverages Docker containers. The source code is MIT licensed and available at https://github.com/cloudve/galaxycloudrunner. The documentation is available at http://gcr.cloudve.org/.ContactEnis Afgan ([email protected])Supplementary informationNone

Download Full-text