wiki:WikiStart
Last modified 2 years ago Last modified on 10/28/15 21:19:19

Welcome to the Research Computing System UABgrid Development Project

Overview

The Research Computing System (RCS) UABgrid Collaboration Environment is a vision for a user controlled collection of applications, tools, and services integrated through a uniform understanding of user identity and the groups (communities) to which the user belongs. This environment enables seamless access to shared resources and facilitates collaborations within labs, across campus and around the globe.

The RCS UABgrid development project is a collaborative effort to build a private cloud programatically controllable suite of services to support the construction and operation of research applications. The RCS UABgrid is the collection of these resources. and the communities surrounding their development.

RCS UABgrid is an open development effort that welcomes participation of those interested in the construction, use, or operation of this platform infrastructure. The development project is supported by the services of the RCS UABgrid Collaboration Environment as a demonstration of its features, as an assertion of confidence in its design, and as a method of ensuring that problems are discovered and resolved effectively. We eat our own dog food.

The computer geek in you may recognize that these themes echo the development of distributed operating systems. The platform strives for neutrality so developers can leverage select components to build the applications that best meet their needs. Applications can be as simple or complex as needed. Follow the plan that's right for you. We will follow Plan9.

Activities

We continually improve the features of the compute, virtualization, and storage capacity to support development of research applications.

We hold a weekly DevOps meeting where we plan our work on this fabric.

Current topics

Feel free to keep a pulse on the operation, support and progress of Research Computing System development by subscribing to an RSS feeds available throughout the project.

Past Work

Dell OpenStack+Ceph PoC April 2013: We have begun a project with Dell and Inktank to deploy their OpenStack-Powered cloud solution on the 192 core, 1TB RAM, and 420TB hardware fabric we acquired in 2012Q4 to provide general purpose storage solutions on the Research Computing System. Follow the development on this exciting addition to our platform that will usher our research storage and virtualization into production in 2013Q22013Q4 (oops). The OpenStackPlusCeph project page sets the stage and will be the information nexus for our solution.

Our initial focus was the expansion of HPC resources through grid computing, and the origin of the name "UABgrid". The world wide HPC community gave rise to the grid infrastructure through the Globus Toolkit and other tools designed to harness resources distributed across many organizations. We continue to develop access to large scale computational resources by leveraging grid interfaces of the Open Science Grid (OSG) and XSEDE.

We have also helped demonstrate the utility of grid technologies for the bio-informatics community by helping the Comprehensive Cancer Center and the Center for Clinical and Translational Science deploy pilot applications as part of the Cancer Biomedical Informatics Grid (caBIG) and the CTSA initiatives. Specifically, the Aggregate Cohort Estimator (ACE) uses technologies from both caBIG and InCommon? to facilitate access to pre-investigation patient cohort estimates.

Research Computing System Components

The Research Computing System is being designed to provide high-performance computing (HPC), storage, and virtual machine resources for the support of research applications. The new storage and virtual machine resources are under active development.

Computation

Storage

  • 180TB HPC Storage -- a DDN-based Lustre fabric for fast access to large files
  • 432TB General Storage -- general storage to support research application development Under Development'

Virtual Resources

  • Virtual machines -- to support research application development Under Development

Collaboration Tools and Applications

The following are actively used examples of resources built using and run on the RCS.

  • Docs - the project managing the documentation resources
  • Projects - the pages for managing the projects instance.
  • Blogs - the project pages for the managing the blogs instance.
  • Galaxy - configuration notes for Galaxy

Documentation

Support

If you have a problem using UABgrid please open a ticket and describe your issue.

Mailing Lists

UABgrid Security Announcements

  • UGS-2008-001 - UABgrid User Certificates impacted by Debian OpenSSL Vulnerability

UABgrid Incident Reports

UABgrid Incident Reports (UGI) are non-security related event reports that related to significant unexpected service interruptions.

Topics of Interest

Resource support pages

  • Cheaha - a 64 node AMD-based compute cluster
  • CoreCluster - The core cluster is the platform that hosts shared infrastructure for the operation and management of UABgrid

General Help