Search

This report describes our work on the Datacomputer system from July 1, 1975 to December 31, 1975. Work during the reporting period falls into two main categories: installation and operation of Datacomputer Version 1, the first full service version of the Datacomputer; and preparation for the next version, which is to incorporate an Ampex Terabit Memory System. Parallel operation of Version 0/11 continued into this period, and various other activities received attention as will be reported.

The purpose of the project is to support the ARPA-NMRO Seismic Data Activity by providing data storage and retrieval services., See also report dated 21 Mar 1975, ADA006932.

The Packet Speech Measurement Facility (PSMF) is an investigative tool designed to be used by researchers in packet network studies. The PSMF facilitates experiments dealing with the timing and composition of packet flow, and will help elucidate the interactions between the conceptual structures of protocol design and the physical exigencies of network implementation. This report summarizes efforts aimed at the generalization of existing PSMF functions to serve a larger community of packet network experimenters, and have culminated in the design specification for the augmented features, the implementation of additional protocols, and the installation of some server facilities.

Project activity can be divided into four catagories: (1) SIP development and network bandwidth considerations; (2) TBM acquisition and integration into the Datacomputer; (3) coordination with the seismic community; and (4) seismic-data related Datacomputer development. One section in this report is devoted to each of these catagories., See also report dated 31 Oct 1975, ADA019961.

During 1978, Computer Corp. of America offered very large on-line data storage and retrieval services on the Datacomputer in support of the seismic community data activities and for general use. The Datacomputer is a system designed to allow convenient and timely access to large on-line databases for multiple remote users communicating over a network. The Datacomputer is only operational general purpose database system capable of handling data sets in excess of a trillion bits. Copious and inexpensive storage is a unique feature of the CCA Datacomputer, made possible by the incorporation of an Ampex Tera-Bit Memory System (TBM). The TBM at CCA was the first public installation of this video-tape technology based system. The CCA installation is configured to hold up to 175 billion bits on-line with four TBM tape drives. Additional data (almost entirely seismic) is stored in off-line TBM tapes.

For a number of years, the Nuclear Monitoring Research Office (NMRO) has sponsored the development of tools and techniques which have been of value in advancing seismological research. But only with the development of large- database techniques has it been possible for seismologists to do meaningful analytical work over the vast quantities of data which represent seismic activity. Instruments, installed world-wide, detect, acquire and measure seismic readings. And computers are employed in virtually every aspect of collecting, analyzing, processing and storing seismic waveform data. At CCA, the Data computer (MARILL and STERN) was developed under ARPA funding to function as a first-of-a-kind network data utility for the support of very large databases, allowing shared remote access by heterogeneous computers in a network environment. The Datacomputer thus provides the facility required to store online the very large amounts of seismic data being collected.

This report describes very large on-line data storage and retrieval services over the Arpanet to support seismic data activity and general use. Use of these services continued to grow in the Arpanet community and additional applications are under active investigation.

Computer Corporation of America continued to provide Datacomputer and SIP service to the ARPAnet community throughout the final quarter of 1979. The Datacomputer is a network-oriented database management utility which is used in diverse general networking applications. The Datacomputer's software system is designed to allow convenient and timely access to large on-line databases and to promote data sharing by multiple remote users communicating over a network. The seismic application was heretofore the largest user of the Datacomputer. This was true not only in terms of the amount of data stored, but also in terms of its complexity and the required bandwidth of transfer. CCA developed and operated a dedicated Seismic Input Processor (SIP) to field the real-time input stream. Both the real-time data stream storage via the SIP and the SRO files storage from the Albuquerque Seismological Laboratory were shut down during this reporting period.

The purpose of the project is to support the ARPA-NMRO Seismic Data Activity by providing data storage and retrieval services. The Arpanet will be used as the communication channel. As part of the service, seismic data will be (a) collected from the Arpanet; (b) stored; and (c) made available to computers on the Arpanet in a convenient and timely manner. These services will represent a special application of the Arpanet Datacomputer., See also AD0787017.

The datacomputer system is being designed as a large-scale data storage utility to be accessed from remote computers on the Arpanet and, potentially, on other networks. The development is phased, with each successive release of the system offering increased capabilities to users. During the present reporting period, the second release of the system became operational. This release, while still primitive in many respects, is beginning to provide experience with actual applications and user programs., See also AD0770881. MDA903-74-C-0225. Sponsored in part by DARPA

The Datacomputer system is being designed as a large-scale data storage utility to be accessed from remote computers on the Arpanet and, potentially, on other networks. The development is phased, with each successive release of the system offering increased capabilities to users. During the present reporting period, the third major release of the system became operational. This release, while still primitive in many respects, is providing service for a wide range of applications., See also report dated Jun 1974, AD0787677.

This report describes Computer Corporation of America's (CCA's) contributions to the U.S. Air Force Strategic Air Command (SAC) Command, Control, and Communication (C3) Experiment. The objective of the Experiment is to demonstrate how emerging communication and software technologies can be integrated to support survivable trans- and post-attack command and control. CCA has provided software research, design, and implementation support in the area of distributed database management technology. This technology will allow systems to provide continuous access to critical information through intelligent, easy-to-use interfaces to replicated and distributed strategic databases. The first subsection below summarizes CCA's contributions to the Experiment. The second subsection describes how CCA's work addressed each point in the project's Statement of Work.

This report contains a collection of papers that detail the design of a SDD-1 (A System for Distributed Databases). It is the second part of the final technical report for a research project called A Distributed Database Management System for Command and Control Applications. Each paper included in this report is a self-contained entity in that the papers can be read in any order without loss of understanding. However, if a reader is not specifically interested in one particular aspect of SDD-1, it is suggested that the first paper be read first. The first paper presents the overall design of SDD-1. The second paper describes the SDD-1 concurrency control algorithm in detail. The third paper is a proof of the correctness of the SDD-1 concurrency control algorithm. The fourth paper describes the reliability mechanisms of SDD-1. The last paper focuses on the distributed query processing technique used by SDD-1. (Author).

This document is the final report on the design and implementation of a program visualization (PV) environment, intended to offer the user an integrated graphics programming support system. The PV environment has capitalized on recent progress in the graphical representation of information, to provide designers and programmers with both static and dynamic (animated) views of systems. The PV research prototype supports programming in C, although large portions of the system are independent of the software development language. The goal of program visualization is to facilitate the understanding of programs by people. To visualize means 'to see or form a mental image of'. Successful program visualizations aid programmers in the formation of clear and correct mental images of the structure and function of programs.

SDD-1, A System for Distributed Databases, is currently being designed and implemented by Computer Corporation of America. SDD-1 is composed of a collection of datamodules which may be dispersed geographically and which are assumed to communicate over channels which may vary in bandwidth and delay. Also, the system supports redundant databases, meaning that portions of databases may be stored at two or more datamodules in order to improve the reliability and efficiency of database operations. This paper presents a brief overview of several key facets of SDD-1. These include: the system architecture, the data distribution concepts used by the system, SDD-1's approach to directory management, the method employed for efficiently handling updates to redundantly stored data, and SDD-1's method for efficiently retrieving data which is dispersed at distant datamodules.

This paper addresses the following problem: how to perform updates in a redundant distributed database system (RDDBS) in a manner that preserves the consistency of the database, yet does not introduce intolerable inter-computer synchronization delays. We have developed a technique for this problem which permits most update transactions to exhibit the same highly responsive behavior which an RDDBS can offer on retrieval. Other approaches to the problem of updating in a redundant distributed database environment suffer from substantially slower response times (in many cases involving clearly intolerable delays) as well as from problems in system scaling.

This paper reports on the design of a program visualization (PV) environment, intended to provide lifecycle support for software development. The PV environment will capitalize on recent progress in the graphical represnetation of information and low-cost color graphics, to provide designers and programmers with both static and dynamic (animated) views of systems. The aim is to support maintainers of large complex sofware systems. The PV system will, in effect 'open the side of the machine' to permit users to look inside and watch their programs run. In this paper, we survey categories of program illustrations and then present and motivate the design philosophy that we are pursuing for the PV environment.

Spatial Data Management is a technique for organizing and retrieving information by positioning it in a Graphical Data Space (GDS). This Graphical Data Space is viewed through a color raster scan display which enables users to traverse the GDS surface or zoom into the image to obtain greater detail. In contrast to conventional database management systems in which users access data by asking questions in a formal query language, a Spatial Data Management System (SDMS) presents the information graphically in a form which seems to encourage browsing and to require less prior knowledge of the contents and organization of the database. (Author), Original contains color plates: All DTIC and NTIS reproductions will be in black and white.

This paper describes the techniques used to optimize relational queries in the SDD-1, distributed database system. Queries are submitted to SDD- 1 in a high-level procedural language called Datalanguage. Optimization begins by translating each Datalanguage query into a relational calculus form called an envelope, which is essentially an aggregate-free QUEL query. This paper is primarily concerned with the optimization of envelopes. Envelopes are processed in two phases. The first phase executes relational operations at various sites of the distributed database in order to delimit a subset of the database that contains all data relevant to the envelope. This subset is called a reduction of the database. The second phase transmits the reduction to one designated site, and the query is executed locally at that site. The critical optimization problem is to perform the reduction phase efficiently. Success depends on designing a good repertoire of operators to use during this phase, and an effective algorithm for deciding which of these operators to use in processing a given envelope against a given database. The principal reduction operator that we employ is called semi-join. In this paper we define the semi-join operator, explain why semi-join is an effective reduction operator, and present an algorithm that constructs a cost effective program of semi-joins given an envelope and a database.

The key objective of this project is to design and implement a distributed database management system called SDD-1 (System for Distributed Databases). SDD-1 is being designed as a cooperating set of processors called datamodules. Each datamodule stores a portion of the database and a given data item may be stored redundantly at several datamodules. A user accessing SDD-1 at any datamodule can retrieve or update data stored anywhere in the network. In fact, the user need not be aware of the distribution at all. A database system with this distributed architecture is important for command and control because it supports access to an integrated database while offering the following key operational advantages: Reliability/survivability; Tunable efficiency; and Modular upward scaling.

Computer Corporation of America (CCA) has completed a one-year project designed to continue and expand the operation of a Demonstration and Development Facility (DDF). The DDF was a first-of-a-kind research facility used to integrate, demonstrate, and transfer computer-based research products. The DDF also provided computer resources and support services for the development of entirely new research products. The DDF began in 1977 under funding from what was then the ARPA Cybernetics Technology Office (CTO). During its first year, the DDF devoted itself exclusively to CTO-funded Crisis Management research. Beginning with fiscal 1979, CCA expanded the scale of the DDF operation to serve researchers working on all CTO programs. CCA also conducted intensive technology transfer efforts in support of four selected CTO research programs: Combat Readiness/Effectiveness, Advanced Decision Technology, Command Systems Cybernetics, and Crisis Management. In the three other selected programs, the activity was new to the DDF but closely related to previous CCA work.

The Crisis Management Development and Demonstration Facility provided five types of services: (1) Timesharing computer services to Crisis Management Program personnel and contractors. (2) Assistance to the Crisis Management Program in selection and acquisition of state-of-the-art technologies for installation in the facility. (3) Assistance to Crisis Management Program researchers in using the Facility, in converting applications software to run on the Facility's computer system, and in computer-system aspects of creating new software products. (4) Demonstration of software products developed at the DDF to potential users of the products. (5) A product distribution service in which developed software products were packaged and distributed to operational users. This report discusses the three phase of the DDF life cycle: Phase I: procurement and installation of a hardware configuration at the site, together with initial software installation. Phase II: Operation of the timesharing system, and the development and demonstration of the database and related software. Phase III: Completion of system software acquisition and the development of application software.

This report describes the USS Carl Vinson application of Computer Corporation of America's (CAA's) Spatial Data Management System (SDMS) -- a unique, user interface that provides a simple and uniform means of accessing different kinds of computerized information. Work under the referenced contract consists of the design and partial implementation of this application. SDMS is a computer-based tool that uses graphical symbols to convey information. SDMS merges the visual power of symbols with the information-handling facilities of a conventional database management system (DBMS). It provides a highly effective means for people who are not data processing professionals to organize and access a database. Spatial data management is a technique for organizing and retrieving information by representing and positioning it graphically. Data are viewed through color displays. The displays show flat data surfaces on which pictorial representations of the data are arranged. The collection of all the data surfaces is called the graphical data space (GDS). The GDS consists of all the pictures that the user can access. SDMS automatically creates these pictures from data stored by the DBMS., Original contains color plates: All DTIC and NTIS reproductions will be in black and white.

This is the final report of a project by Computer Corporation of America (CCA) to design and implement a prototype Spatial Data Management System (SDMS). SDMS is a unique, man-machine interface that provides a uniform means of accessing different kinds of computerized information., See also report dated 31 Aug 1979, ADA083948.

Navy acquisition managers can be provided with improved information and better decision support tools than are currently available to them through the application of recent advances in information technology. Of several acquisition management areas that could benefit from that technology, acquisition planning has the potential for the most immediate and important gains. This conclusion is the result of analysis by ROH Incorporated and Computer Corporation of America, a team with expertise in both Navy acquisition practice and relevant computer technology. The team interviewed managers in NAVSEA and NAVMAT, collected and studied an extensive library of relevant documents, and employed a recognized system analysis methodology in reaching this conclusion. This report presents that analysis of acquisition management activity, supports and details its key conclusions, and defines the scope of the necessary work to follow. The key recommendation is that the Navy continue with a detailed study of acquisition planning and related activity., Prepared in cooperation with ROH, Inc., Arlington, VA, on contract N00014-79-C-0928.

This report summarizes a project entitled 'The Enhancement of Datamodule 1.' The focus of this project is to study the Datacomputer's (MARILL and STERN) performance in its role as a datamodule in SDD-1 and as a DBMS in other command and control applications. The goal of this study is to produce a set of potential performance enhancements and an analysis of their expected effect on overall Datacomputer performance. Some of the proposed enhancements would be subsequently implemented in the Datacomputer. During the reporting period, the model of Datacomputer performance constructed previously during this project was used to analyze performance in the Advanced Command and Control Architecture Testbed (ACCAT). The Datacomputer was instrumented to produce extensive performance data at runtime. The results of this work have indicated which parts of the system are bottlenecks and this, in turn, is suggesting potential enhancements. In addition, the instrumented Datacomputer has been used in other performance work for ARPA.

This report describes the architecture of the Multibase system. Details of the components of the architecture to be implemented in the initial breadboard version are also described. Although additional research is requierd to fill in the details of optimization and incompatible data handling, the architecture already contains several innovative ideas in integrating distributed heterogeneous databases. These include: the idea of using an integration data base to resolve data incompatibility; the idea of using a mapping language to uniformly define the global schema in terms of the local schemata and the integration schema; and the idea of using query modification and query graph decomposition to transform a global query into local queries and queries over the integration database. (Author)

This technical report describes work completed on a prototype Spatial Data Management System (SDMS). The Graphical Display of Data presents an overview of the icon creation mechanisms currently included in SDMS. Icon creation is the facility by which graphical representations of symbolic data are automatically produced. These graphical representations are referred to as 'icons'. Section two includes detailed discussions of the icon creation features added to the system: Secondary data a mechanism by which information in multiple relations or tuples may be presented in a single icon, and a subicons mechanism that allows the appearance of a portion of an icon to be described independently of its parent icon.

This report summarizes the fifth six month period of a project entitled: 'A Distributed Database Management System for Command and Control Applications'. The primary focus of this effort is to design and implement a distributed database management system called SDD-1 (System for Distributed Databases). SDD-1 being installed in phases into the Advanced Command and Control Architectural Testbed (ACCAT) at the Naval Ocean Systems Center (NOSC). SDD-1 is a database management system (abbr. DBMS) with capabilities similar to those found in many currently available centralized DBMSs. However, unlike any currently available DBMS, SDD-1 is implemented on multiple processors geographically distributed and interconnected in a computer network.

The Packet Speech Measurement Facility (PSMF) is an investigative tool designed to be used by researchers in packet network studies. The PSMF facilitates experiments dealing with the timing and composition of packet flow, and will help elucidate the interactions between the conceptual structures of protocol design and the physical exigencies of network implementation. This report summarizes efforts undertaken by the Computer Corporation of America during the second year of PSMF development: January 1, 1978 to December 31, 1978. These efforts have been aimed at the generalization of existing PSMF functions to serve a larger community of packet network experimenters, and have culminated in the design specification for the augmented features, the implementation of additional protocols, and the installation of some server facilities., See also ADA061346.

During 1978, Computer Corporation of America assisted the Vela Seismological Center (VSC) and the Seismic Data Analysis Center (SDAC) in the design, creation, and management of the Signal Waveform Files (SWF). The SWF system is intended to make all of the recent 'interesting' seismic data of the Vela Seismological Network available in the Datacomputer. This work is continuing in 1979 under a follow on proposal.

This fourth quarter of work on the design and implementation of a prototype Spatial Data Management System (SDMS) resulted in the first operational version of the system. The bulk of this effort was devoted to constructing the mechanism through which a user can create, modify, and view an image plane. An image plane is a flat surface upon which a user can store information. As that surface may be considerably larger than the display screen, the SDMS provides a window which can be moved over that surface in order to view parts of it. As the window is moved, the window's position is indicated on an auxiliary screen which serves as a navigational aid containing a map of the entire image plane. This implementation of SDMS at CCA is the first such system to integrate the I-Space creation and viewing operations so that a user can create and modify a database without the need of a skilled computer professional., Original contains color plates: All DDC reproductions will be in black and white.

The SIP is a dedicated minicomputer communications system developed and operated by Computer Corporation of America. It interfaces a real time seismic information network to the Datacomputer (DORIN and EASTLAKE). The Datacomputer is a network data utility designed to support very large databases and shared access by remote heterogeneous computers (MARTLL and STERN). It is the only operational general purpose database system capable of handling datasets in excess of a trilliom bits (CCA).

This document contains listings of the SWF-D program source-code and data definitions. The source-code is divided here into sections, each of which represents a separately compiled program module and, to facilitate referencing, subsection numbers have been assigned to several of the more interesting routines within the modules. Global data definitions, constants and variables are listed in the Appendix.

Spatial Data Management is a technique for organizing and retrieving information by positioning it in a Graphical Data Space. In contrast to conventional database management systems (DBMSs) which require that information be stored and retrieved by specifying attributes as numbers or strings of text, Spatial Data Management allows a user to employ the spatial location and visual appearance of information in order to find it. The underlying concept is that spatial location is, for many purposes, easier to remember and work with than the keywords of an ordinary DBMS. By allowing a datum to be stored in proximity to related pieces of information, a Spatial Data Management System (SDMS) frees the user of the need to know the exact name or location of the information that he seeks. Instead, he can locate it by 'browsing' until he finds something that he can identify visually. Spatial management of data can be combined with conventional DBMS techniques to yield a system which provides both means of access to information.

The seismic data activity sponsored by the Nuclear Monitoring Research Office (NMRO) involves the collection, storage and processing of seismic waveform information as measured by instruments installed throughout the world. The data activity is intended to assist seismologists in exploring techniques for the detection of seismic events, for pinpointing their location, and for recognizing the causes of these events.

This report describes the work on the Datacomputer, a network data utility, from January 1, 1975 to June 30, 1975. The work is described in detail in sections 2-8. Section 2 is a discussion of the Datacomputer architecture, with emphasis on the increasing levels of functional abstraction beginning with the hardware and moving outward. Section 3 is a report on the usage of the Datacomputer during the reporting period, and a discussion of new work being done in the user services and support area. Section 4 is a detailed discussion of the work on the Datacomputer software carried out during the period under discussion. Most of the effort was concentrated in this area. Section 5 discusses the on-going work of documenting the Datacomputer. Section 6 describes progress made in the area of Datacomputer hardware and operational support. Section 7 is a bried overview of the NMRO work and its implications for the Datacomputer in general. Finally, Section 8 is a catch-all for minor but important areas of Datacomputer development., See also report dated 31 Dec 1974, ADA008877.

This is the second of three volumes of the final technical report for the project Distributed Database Control and Allocation. This volume describes work on the performance analysis of concurrency control algorithms., See also Volume 3, AD-A138 893.

This is the third of three volumes of the final technical report for the project Distributed Database Control and Allocation. This volume attempts to provide a handbook of information about a number of important concurrency control algorithms which can be used in the design of a distributed data base management system., See also Volume 1, AD-A138 891.

This is the first of the three volumes of the final technical report for the project Distributed Database Control and Allocation. This volume describes frameworks for understanding concurrency control and recovery algorithms for centralized and distributed database systems., See also Volume 2, AD-A138 892.

The focus of this project is to study the Datacomputer's performance in its role as a datamodule in SDD-1 ROTHNIE and GOODMAN and as a DBMS in other command and control applications. The goal of this study is to produce a set of potential performance enhancements and an analysis of their expected effect on overall Datacomputer performance. During the reporting period, a model of Datacomputer performance was constructed and measurements of current systems performance were made. The results of this initial work have already indicated that parts of the system are bottlenecks and this, in turn, is suggesting potential enhancements to be analyzed further along in the study.

The Packet Speech Measurement Facility (PSMF) is an investigative tool designed to be used by researchers in packet network studies. The PSMF facilities experiments dealing with the timing and composition of packet flow, and will help elucidate the interactions between the conceptual structures of protocol design and the physical exigencies of network implementation. This report summaries efforts aimed at the generalization of existing PSMF functions to serve a larger community of packet network experimenters, and which have culminated in the design specification for the augmented features, the implementation of additional protocols, and the installation of some server facilities.

SDD-1, a System for Distributed Databases, is a distributed database system being developed by CCA. SDD-1 permits data to be stored redundantly at several database sites in order to enhance the reliability and responsiveness of the system and to facilitate upwards scaling of system capacity. This paper describes the algorithm used by SDD-1 for updating that is stored redundantly.