Gold is an open source accounting system that tracks and manages resource usage on High Performance Computers. It acts much like a bank in which resource credits are deposited into accounts with access controls designating which users, projects, and machines may access the account. As jobs complete or as resources are utilized, accounts are charged and resource usage recorded. Gold supports familiar operations such as deposits, withdrawals, transfers, and refunds. It provides balance and usage feedback to users, managers, and system administrators.
Since accounting needs vary widely from organization to organization, Gold has been designed to be extremely flexible, featuring customizable accounting and supporting a variety of accounting models. Attention has been given to scalability, security, and fault tolerance. Gold facilitates the sharing of resources between organizations or within a Grid by providing distributed accounting while preserving local site autonomy.
Gold is being developed at Pacific Northwest National Laboratory (PNNL) as open source software under the Scalable Systems Software (SSS) SciDAC project. Gold is currently in alpha release and is beginning alpha testing at a number of DOE and university sites.
Gold was designed to meet the accounting needs of computing centers that share resources in multi-project environments. In order for an organization to use its high performance computers most effectively, it must be able to allocate resources to the users and projects that need them in a manner that is fair and according to mission objectives. Tracking the historical resource usage allows for insightful capacity planning and in making decisions on how to best mete out these resources. It allows the funding sources that have invested heavily in a supercomputing resource a means to show that it is being utilized efficiently.
Gold was also designed to facilitate the sharing of resources between organizations or within a Grid to take advantage of the tremendous utilization gains afforded by meta-scheduling.
|© 2001-2010 Adaptive Computing Enterprises, Inc.|