Publications

Internet Protocol version 4 (IPv4) has been a mainstay of the both the Internet and corporate networks for delivering network packets to the desired destination. However, rapid proliferation of network appliances, evolution of corporate networks, and the expanding Internet has begun to stress the limitations of the protocol. Internet Protocol version 6 (IPv6) is the replacement protocol that overcomes the constraints of IPv4. As the emerging Internet network protocol, SNL needs to prepare for its eventual deployment in international, national, customer, and local networks. Additionally, the United States Office of Management and Budget has mandated that IPv6 deployment in government network backbones occurs by 2008. This paper explores the readiness of the Sandia National Laboratories network backbone to support IPv6, the issues that must be addressed before a deployment begins, and recommends the next steps to take to comply with government mandates. The paper describes a joint work effort of the Sandia National Laboratories ASC WAN project team and members of the System Analysis & Trouble Resolution, the Communication & Network Systems, and Network System Design & Implementation Departments.

Abstract not provided.

Abstract not provided.

In recent years, modeling and simulation has played an increasingly important role in the maintenance of the nuclear stockpile. The Advanced Simulation and Computing (ASC) program continues to support and encourage the development of a modeling and simulation infrastructure to make these goals a reality. The Distance Computing Network has been making make the ASC resources available to users throughout the tri-lab environment for over five years. This network relies on the Transmission Control Protocol/Internet Protocol (TCP/IP) protocol suite to provide high performance and reliable communications. Understanding TCP/IP operation in this unique environment is critical. Software modeling has been used to analyze current network performance and predict the effect of proposed changes. Recently the network architecture was radically changed and the software model had to be changed as well. Whereas the original network was based on 2.5 gigabit per second ATM links, the redesigned network is comprised of 10-gigabit Ethernet links arranged as a 3-node ring. Therefore, a new software model was needed to continue to predict the performance of proposed changes and allow engineers to experiment with new network applications without the risk of interfering with critical operations.

Internet Protocol version 4 (IPv4) has been a mainstay of the both the Internet and corporate networks for delivering network packets to the desired destination. However, rapid proliferation of network appliances, evolution of corporate networks, and the expanding Internet has begun to stress the limitations of the protocol. Internet Protocol version 6 (IPv6) is the replacement protocol that overcomes the constraints of IPv4. IPv6 deployment in government network backbones has been mandated to occur by 2008. This paper explores the readiness of the Sandia National Laboratories' network backbone to support IPv6, the issues that must be addressed before a deployment begins, and recommends the next steps to take to comply with government mandates. The paper describes a joint, work effort of the Sandia National Laboratories ASC WAN project team and members of the System Analysis & Trouble Resolution and Network System Design & Implementation Departments.

The Advanced Simulation and Computing (ASC) Distance Computing (DisCom) Wide Area Network (WAN) is a high performance, long distance network environment that is based on the ubiquitous TCP/IP protocol set. However, the Transmission Control Protocol (TCP) and the algorithms that govern its operation were defined almost two decades ago for a network environment vastly different from the DisCom WAN. In this paper we explore and evaluate possible modifications to TCP that purport to improve TCP performance in environments like the DisCom WAN. We also examine a much newer protocol, SCTP (Stream Control Transmission Protocol) that claims to provide reliable network transport while also implementing multi-streaming, multi-homing capabilities that are appealing in the DisCom high performance network environment. We provide performance comparisons and recommendations for continued development that will lead to network communications protocol implementations capable of supporting the coming ASC Petaflop computing environments.

Policy-based network management (PBNM) uses policy-driven automation to manage complex enterprise and service provider networks. Such management is strongly supported by industry standards, state of the art technologies and vendor product offerings. We present a case for the use of PBNM and related technologies for end-to-end service delivery. We provide a definition of PBNM terms, a discussion of how such management should function and the current state of the industry. We include recommendations for continued work that would allow for PBNM to be put in place over the next five years in the unclassified environment.

The desire to provide a measure of computer network availability at Sandia National Laboratories has existed for along time. Several attempts were made to build this measure by accurately recording network failures, identifying the type of network element involved, the root cause of the problem, and the time to repair the fault. Recognizing the limitations of available methods, it became obvious that another approach of determining network availability had to be defined. The chosen concept involved the periodic sampling of network services and applications from various network locations. A measure of ''network'' availability was then calculated based on the ratio of polling success to failure. The effort required to gather the information and produce a useful metric is not prohibitive and the information gained has verified long held feelings regarding network performance with real data.

Windows NT desktop and server systems are becoming increasingly important to Sandia. These systems are capable of network performance considerably in excess of the 10 Mbps Ethernet data rate. As alternatives to conventional Ethernet, 155 Mbps Asynchronous Transfer Mode, ATM, and 100 Mbps Ethernet network interface cards were tested and compared to conventional 10 Mbps Ethernet cards in a typical Windows NT system. The results of the tests were analyzed and compared to show the advantages of the alternative technologies. Both 155 Mbps ATM and 100 Mbps Ethernet offer significant performance improvements over conventional 10 Mbps shared media Ethernet.

Computer networks, supporting an organization's activities, are prevalent and very important to the organization's mission. Implementing a heterogenous organizational network allows the staff to select the computing environment that best supports their job requirements. This paper outlines the lessons learned implementing a heterogenous computer network based on networking standards such as TCP/IP and Ethernet. Such a network is a viable alternative to a proprietary, vendor supported network and can provide all the functionality customers expect in a computer network. 2 figs.

Sandia National Laboratories, Albuquerque, may soon have more responsibility for the operation of its own telephone system. The processes that constitute providing telephone service can all be improved through the use of a central data information system. We studied these processes, determined the requirements for a database system, then designed the first stages of a system that meets our needs for work order handling, trouble reporting, and ISDN hardware assignments. The design was based on an extensive set of applications that have been used for five years to manage the Sandia secure data network. The system utilizes an Ingres database management system and is programmed using the Application-By-Forms tools.