Big Data analytics using Spark, Hadoop and SparkR can be conducted on XSEDE Bridges. In order to use the Bridges supercomputer, students, researchers, faculty and staff members are required to attain an allocation of Service Units (SUs). Service Units can be traded for time on Bridges.
If you are just getting started, ITS Research Services can provide a small number of Service Units through the XSEDE Campus Champions Program for you to test the system capabilities before you officially request allocations from XSEDE. There are three official allocation types:
Startup Allocations - Requesting a startup allocation is a great way to begin using big data analytics tools. Appropriate uses for Startup Allocations include:
- Application development by researchers and research teams
- Benchmarking, evaluation and experimentation on the various resources
- Testing existing data analytics workflows
Education Allocations - This flavor of Service Unit Allocation is reserved for academic courses or training activities that have specific start and end dates. Examples of appropriate uses for Education Allocations include:
- Conducting a class that utilizes Spark/Hadoop/SparkR
- Conducting a training session or an introductory seminar on Big Data Analytics.
Faculty that wish to conduct a course that utilizes Spark, Hadoop or SparkR can request Service Unit allocations through the Education Allocation Requests process outlined here. To use XSEDE Bridges in your research, a Research Allocation of Service Units must be applied for. More information on Research Allocations can be found on the XSEDE website.
Research Allocations - Once users have progressed beyond the startup phase and gained insight into their workflow, researchers can officially request a Research Allocation. Research requests are accepted and reviewed quarterly by the XSEDE Resource Allocations Committee (XRAC).
Access to all XSEDE platforms is conducted through the XSEDE single sign-on hub. Click here for information on signing up and accessing Bridges.