High-End Services Section (HSS)  
     
  CISL Supercomputer Services Group - We manage the compute engines that drive NCAR science.  
     
   
advanced  
 

 
 Supercomputer Services Group - Projects

SSG Projects

Here you will find information about the current SSG projects.

[ High Priority | Medium Priority | Ongoing ]
High Priority
Title: Centralized Power Up/Down Console Deplolyment
Purpose: The purpose of this project is to develop a centralized power up/down system console facility for CPG.
Status: Initial testing is underway.
Last Updated: October 02, 2006
[Top]
Title: File System Scrubber Rearchitecture
Purpose: The purpose of this project is to develop and deploy a new file system scrubber methodology across all of the supercomputers.
Status: Work is currently underway to repair the existing scrubber in preparation for the blueice rearchitecture phase of the project.
Last Updated: October 02, 2006
[Top]
Title: GPFS MultiCluster
Purpose: The purpose of this project is to develop and deploy a GPFS MultiCluster SAN to support the supercomputer divisional file systems.
Status: Nothing new to report at this time.
Last Updated: October 02, 2006
[Top]
Title: ICESS Installation, Configuration, & Deployment
Purpose: The purpose of this project is to take the new ICESS system from selection to production.
Status: Installation planning is now underway
Last Updated: October 02, 2006
[Top]
Title: SMART Development & Deployment
Purpose: The purpose of this project is to develop and deploy the System Management And Reporting Tool (SMART).
Status: Nothing new to report at this time.
Last Updated: October 02, 2006
[Top]
Title: SSG Central License Server Replacement
Purpose: The purpose of this project is to replace the chickenhawk license server with newer technology hardware and software.
Status: Nothing new to report at this time.
Last Updated: October 02, 2006
[Top]
Title: Snoopy HSS Data Repository
Purpose: Install, configure, and deploy snoopy as the HSS metrics data repository.
Status: No new status to report at this time.
Last Updated: October 02, 2006
[Top]
Title: Special Computing Campaign Support
Purpose: The purpose of this project is to support the ongoing series of special computing campaigns on the supercomputers.
Status: The MMM Annual Hurricane campaign began on July 1, 2006 and runs until October 31, 2006. Components are running on lightning and bluevista.
Last Updated: October 02, 2006
[Top]
Medium Priority
Title: Batch System Data Mining & Simulator Project
Purpose: The purpose of this project is to begin to mine the LSF accounting data in search of scheduler optimizations and to examine the feasibility of developing a simulator for the NCAR supercomputer environment.
Status: The data mining phase of the project is now underway.
Last Updated: October 02, 2006
[Top]
Title: CISL Resource Information System (CRIS) Support
Purpose: The purpose of this project is to provide design and implementation input into the CRIS database design and business logic re-engineering project.
Status: We are currently awaiting the Phase II architecture document.
Last Updated: October 02, 2006
[Top]
Title: Computational Noise Investigation Project
Purpose: The purpose of this project is to install the ANL computational noise benchmarking tool suite and to eliminate the extraneous sources of computational noise on the NCAR supercomputers.
Status: Nothing new to report at this time.
Last Updated: October 02, 2006
[Top]
Title: IMAGe Cluster Support
Purpose: The purpose of this project is to provide ongoing support for the IMAGe coral cluster and derivative projects.
Status: Nothing new to report at this time.
Last Updated: October 02, 2006
[Top]
Title: LSF Infrastructure Management
Purpose: The purpose of this project is to manage the LSF intrastructure.
Status: LSF Version 7.0 is due out later this year. SSG is working with Platform Computing to become a beta test site for this new version. The early release software will be installed on otis and sparky in late fall with a planned production release in early winter.
Last Updated: October 02, 2006
[Top]
Title: Remote Direct Memory Access (RDMA) Project
Purpose: The purpose of this project is to implement the Remote Direct Memory Access (RDMA) feature on the IBM POWER5 platforms (bluevista, blueice, etc.)
Status: SSG and CSG are working with Platform Computing and IBM to resolve software issues at this time.
Last Updated: October 02, 2006
[Top]
Title: Roy Upgrade & Rearchitecture Support
Purpose: The purpose of this project is to support the DSG roy upgrades and to work with NETS to optimize the supercomputing routing architecture.
Status: No new status to report at this time.
Last Updated: October 02, 2006
[Top]
Title: SSG Security Redesign Project
Purpose: The purpose of this project is to rearchitect the SSG security strategy in parallel with the sister projects in CISL and UCAR.
Status: Nothing new to report at this time.
Last Updated: October 02, 2006
[Top]
Title: SSG System Monitoring Tool Deployment
Purpose: The purpose of this project is to deploy a common set of monitoring tools such as Big Brother and/or Ganglia across the supercomputer infrastructure.
Status: Nothing new to report at this time.
Last Updated: October 02, 2006
[Top]
Title: SSG Web Development & Ongoing Maintenance
Purpose: The purpose of this project is to continually redesign, redeploy, and rearchitect the SSG web site.
Status: The site is now template-based providing a common look and feel. It has been rearchitected to use a combination of static html pages, java server pages, and CISL VAVOOM technologiy. Work is now underway to add additional dynamic content.
Last Updated: October 02, 2006
[Top]
Title: Supercomputer Storage Rearchitecture Project
Purpose: The purpose of this project is to evaluate current NCAR supercomputer storage management policies, procedures, and system architecture and make recommendations for modifications, optimizations, and enhancements.
Status: The GPFS Multicluster project is a child project to this one.
Last Updated: October 02, 2006
[Top]
Ongoing
Title: Batch Job Scheduling Steering Committee
Purpose: The purpose of this project is to periodically review NCAR scheduling policies and practices in search of further optimizations and simplifications.
Status: The mechanism to switch from implicit project declarations to explicit project declarations will be implemented in the October/November timeframe.
Last Updated: October 02, 2006
[Top]
Title: Day-To-Day System Activities
Purpose: The purpose of this project is to handle the myriad day-to-day issues, questions, and problems that arise with regards to the systems, software, tools, and documentation that SSG supports.
Status: SSG is currently working to isolate and resolve the root cause of the lightning login node crashes.
Last Updated: October 02, 2006
[Top]
Title: Maintain Third-Party Software Repository
Purpose: The purpose of this ongoing project is to maintain the third-party software (e.g., GNU products) repository on the NCAR supercomputers.
Status: Nothing new to report at this time.
Last Updated: October 02, 2006
[Top]
Title: Supercomputer System Documentation
Purpose: The purpose of this project is to continue to enhance the supercomputer documentation repository with new and updated content.
Status: Nothing new to report at this time.
Last Updated: October 02, 2006
[Top]
Title: System Accounting Maintenance & Enhancement
Purpose: The purpose of this project is to provide ongoing support for the existing system accounting software and to incorporate new LSF features as they become available.
Status: LSF Version 7.0 provides new system accounting features that SSG will begin to incorporate into the NCAR system accounting process flow early next year.
Last Updated: October 02, 2006
[Top]