skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: A success-history based learning procedure to optimize server throughput in large distributed control systems

Authors:
; ;
Publication Date:
Research Org.:
Brookhaven National Laboratory (BNL), Upton, NY (United States)
Sponsoring Org.:
USDOE Office of Science (SC), Nuclear Physics (NP) (SC-26)
OSTI Identifier:
1405927
Report Number(s):
BNL-113769-2017-CP
R&D Project: KBCH139; 18070; KB0202011
DOE Contract Number:
SC0012704
Resource Type:
Conference
Resource Relation:
Conference: 16th International Conference on Accelerator and Large Experimental Physics Control Systems (ICALEPCS 2017); Barcelona, Spain; 20171008 through 20171013
Country of Publication:
United States
Language:
English
Subject:
43 PARTICLE ACCELERATORS

Citation Formats

Gao Y., Chen, J., Robertazzi, T., and Brown, K. A. A success-history based learning procedure to optimize server throughput in large distributed control systems. United States: N. p., 2017. Web.
Gao Y., Chen, J., Robertazzi, T., & Brown, K. A. A success-history based learning procedure to optimize server throughput in large distributed control systems. United States.
Gao Y., Chen, J., Robertazzi, T., and Brown, K. A. 2017. "A success-history based learning procedure to optimize server throughput in large distributed control systems". United States. doi:. https://www.osti.gov/servlets/purl/1405927.
@article{osti_1405927,
title = {A success-history based learning procedure to optimize server throughput in large distributed control systems},
author = {Gao Y. and Chen, J., Robertazzi, T. and Brown, K. A.},
abstractNote = {},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = 2017,
month =
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:
  • Over the last ten years the development of PC's and workstations has changed the way computing is performed. Previously, extensive computations were performed on large high speed mainframe machines with substantial storage capacity. Large capital and operational costs were associated with these machines. The advent of more powerful workstations has brought more computational cycles to the users at lower cost than was achieved with busy timesharing systems. However, many users still can't afford individual special purpose hardware or gigabytes of storage. A successful distributed processing environment must share these resources. Client/Server models have been proposed to address the issues ofmore » shared resources. They are not a new idea, but their implementation has been difficult. With the introduction of SUN's public domain Remote Procedure Call (RPC) Protocol and SUN's interface generator, RPCGEN, their implementation has been made easier. SUN has developed a set of C'' callable routines that handle the Client/Server operations. The availability of Network File System (NFS) on the SRL CRAY and the arrival of Wollongong's latest version of NFS has allowed applications and information sharing between computing platforms. This paper reviews the Client/Server model with respect to SUN's RPC implementation. The discussion will focus on the RPC connection between local and remote machines, the RPC Paradigm for making remote procedure calls, and the programming levels of the RPC libraries. The paper will conclude with summaries of two applications developed at SRL using the protocol and their effect on our computing environment. These include the Nuclear Plant Analyzer and an animation of fluids using behavioral simulation of atom-like particles.« less
  • In this paper an optimization-based method of drift prevention is presented for learning control of underdetermined linear and weakly nonlinear time-varying dynamic systems. By defining a fictitious cost function and the associated model-based sub-optimality conditions, a new set of equations results, whose solution is unique, thus preventing large drifts from the initial input. Moreover, in the limiting case where the modeling error approaches zero, the input that the proposed method converges to is the unique feasible (zero error) input that minimizes the fictitious cost function, in the linear case, and locally minimizes it in the (weakly) nonlinear case. Otherwise, undermore » mild restrictions on the modeling error, the method converges to a feasible sub-optimal input.« less
  • The authors describe the design and implementation of a distributed parallel storage system that uses high-speed ATM networks as a key element of the architecture. Other elements include a collection of network-based disk block servers, and an associated name server that provides some file system functionality. The implementation is based on user level software that runs on UNIX workstations. Both the architecture and the implementation are intended to provide for easy and economical scalability. This approach has yielded a data source that scales economically to very high speed. Target applications include online storage for both very large images and videomore » sequences. This paper describes the architecture, and explores the performance issues of the current implementation.« less
  • Today`s power plants are no longer controlled in an uncoordinated manner by multiple individual control systems. Instead, modern power plants are controlled by a single unified Workstation based Distributed Control and Information System (DCS). This paper describes control system unification and coordination provides power plants with multiple benefits required for survival in today`s competitive marketplace. Some unification benefits are: (1) Advanced control techniques. (2) Open communication between all portions of the process. (3) Unified plant control -- control of the total generation process; not just the separate control of many individual sub processes. (4) Simplification of training, service, spare partsmore » through the use of one control and information system. (5) High resolution automated graphics providing the ability to zoom into any portion of the plant to obtain real time information and control. (6) Generation of meaningful trends and reports. The use of one common distributed control system including advanced controls and the higher computing power of workstations can provide competitive advantages through reduced operating costs, higher availability, and better operating efficiencies.« less