A formalism for comparing the average execution time of distributed protocols is provided. the comparisons are made independently of the properties of the network on which the protocols are executed. the formalism tak...
详细信息
A formalism for comparing the average execution time of distributed protocols is provided. the comparisons are made independently of the properties of the network on which the protocols are executed. the formalism takes into account computation time, the time to transfer information, the time spent by a site waiting to synchronize with other sites, and the overlap among them. A framework in which the information transfer and synchronization requirements of a protocol are separately and explicitly specified is developed. A knowledge formalism is used to specify the protocol's specification requirements. Transformations on protocols which may change the synchronization structure, the information transferred, or the amount of local computation are defined. It is shown that, if a sequence of such transformations can be applied to a protocol to obtain another protocol, the final protocol runs at least as fast as the initial. Two notions of comparison, containment and reducibility, are given, and their properties are explored. Several protocols, including those for atomic commitment and snapshot recording, are analyzed to illustrate the technique.< >
A comprehensive analysis of file access behavior of several commercial production VAX/VMS environments, based on detailed I/O traces, is presented. the characterization focuses on file control operations (such as open...
详细信息
A comprehensive analysis of file access behavior of several commercial production VAX/VMS environments, based on detailed I/O traces, is presented. the characterization focuses on file control operations (such as opens and closes) which are often expensive in both processing and I/O requirements. the motivation for this analysis is the need to design good data management algorithms for distributed file systems. these algorithms are particularly important for managing client caches. Most of the operations for managing client caches are performed on file opens and closes. the authors present quantitative data on the number of file opens and closes, their timing characteristics, and their declared intent to allow sharing or have exclusive access. the precise dynamic sharing of files is examined.< >
In replicated database systems, a replication control protocol is needed to ensure one-copy serializability. the author incorporates the concept of a regeneration into the missing-partition dynamic voting scheme to de...
详细信息
In replicated database systems, a replication control protocol is needed to ensure one-copy serializability. the author incorporates the concept of a regeneration into the missing-partition dynamic voting scheme to design a replication control protocol. Like the original missing-partition dynamic voting scheme, this protocol supports an inexpensive read operation which accesses one copy, rather than all copies, of each data item read. By incorporating the concept of regeneration and keeping multiple versions for each data item in the database, higher data availability is maintained. To support data regeneration, a replicated directory architecture for the proposed replication control protocol is designed, and it not only supports regeneration of replicated data items, but also provides inexpensive, high-availability directory services, which help maintain database availability.< >
the development of pre-run-time schedulers for certain hard real-time applications is investigated. this scheduling scheme requires the modeling of the program as a set of processes with corresponding timing specifica...
详细信息
the development of pre-run-time schedulers for certain hard real-time applications is investigated. this scheduling scheme requires the modeling of the program as a set of processes with corresponding timing specifications, such as release times, computation times, and deadlines. the models must specify precedence and exclusion constraints between pairs of processes. It is processed by an offline pre-run-time scheduling algorithm that computes a schedule for use at run time, ensuring that all static timing constraints of the system are observed. the technique works best in cases in which most processes are periodic, which is true for the F-18. A modeling process for the M18 Mission Computer software that is designed to lead the application of pre-run-time scheduling to the software and is applicable to other such hard systems is described.< >
A model for programming parallel and distributed computers in which the user is presented with a single linear address space, unlimited computational parallelism, and a global clock for synchronization is presented. T...
详细信息
A model for programming parallel and distributed computers in which the user is presented with a single linear address space, unlimited computational parallelism, and a global clock for synchronization is presented. this allows the user to express the abstract parallelism, without being tied to the target architecture and makes the resulting software transportable. An implementation of this model using C as the source language supplemented with a distributed data storage class and concurrent assignments is presented. An example of generating code for low-level image processing algorithms, including histogram equalization, is given. this example shows that parallel algorithms can be developed without architectural knowledge and the resulting programs can be compiled for a wide range of target processors.< >
Calculations are presented for the electrostatic potential in and around a composite dielectric sphere, containing a noncentral spherical dielectric inclusion, subjected to an axial uniform external field. Some numeri...
详细信息
Calculations are presented for the electrostatic potential in and around a composite dielectric sphere, containing a noncentral spherical dielectric inclusion, subjected to an axial uniform external field. Some numerical results are presented and their significance is discussed. It is shown that the field at the center of the inclusion is only weakly dependent on the degree of eccentricity.< >
Regularization has become an important tool for solving many ill-posed problems in approximation theory-for example, in computer vision-including surface reconstruction, optical flow, and shape from shading. the autho...
详细信息
Regularization has become an important tool for solving many ill-posed problems in approximation theory-for example, in computer vision-including surface reconstruction, optical flow, and shape from shading. the authors attempt to determine whether the approach taken in regularization is always the correct one, and to what extent the results of regularization are reliable. they consider as an example a case in which regularization has been used to reconstruct a surface from sparse data, and attempt to determine how strongly the height of the surface at a certain point can be relied upon. these questions are answered by defining a probability distribution on the class of surfaces considered, and computing its expectation and variance. the variance can be used, for instance, to construct a safety strip around the interpolated surface that should not be entered if collision withthe surface is to be avoided.< >
A pulsed electrical breakdown in n-hexane under quasi-uniform field with E approximately 0.9-2.1 MV/cm has been studied using high-speed laser schlieren recording, statistical analysis, and the effect of external pres...
详细信息
A pulsed electrical breakdown in n-hexane under quasi-uniform field with E approximately 0.9-2.1 MV/cm has been studied using high-speed laser schlieren recording, statistical analysis, and the effect of external pressure. three mechanisms of this breakdown differing in their structures are shown to exist. One corresponds to the electric discharge from a cathode and the two others are due to that from an anode. the pressure effect on the breakdown mechanism is radically different. the analysis of experimental data shows that the occurrence of a certain critical time period and a certain critical interelectrode distance is due to the transition from the cathode breakdown mechanism (bubble mechanism) to the anode discharge involving elements of both bubble and ionization breakdown mechanisms. In the nanosecond range at increased pressures, the ionization mechanism of anode breakdown predominates.< >
WIDE (Widely Integrated distributed Environment) is a research project aimed at achieving a transparent distributed environment over heterogeneous distributedcomputing elements withthe consideration of various types...
详细信息
WIDE (Widely Integrated distributed Environment) is a research project aimed at achieving a transparent distributed environment over heterogeneous distributedcomputing elements withthe consideration of various types of connections for internetworking. the target environment of the research is the computing environment in the academic and research communities, especially in Japan. In order to establish a working environment for research, the project started the WIDE Internet, which provides a backbone of connectivity for the member organizations. the WIDE Internet is connected to research networks in the USA, Australia, and New Zealand via the PACCOM Internet. the primary protocol employed in the project is the Internet Protocol (IP) family. thus, the WIDE Internet provides Japan's IP connectivity for both domestic and international connections. A discussion is presented of technologies being employed for the achievement of project goals such as networking technologies, operating system architecture and its backbone structure, together withthe actual state and the future plans of the WIDE project and WIDE Internet.
LANVista, a distributed analyzer architecture for real-time monitoring of local area networks (LANs), is described. LAN installations range from tens of nodes interconnected over a single homogeneous network, to hundr...
详细信息
LANVista, a distributed analyzer architecture for real-time monitoring of local area networks (LANs), is described. LAN installations range from tens of nodes interconnected over a single homogeneous network, to hundreds of nodes distributed on a variety of network types interconnected through repeaters, bridges and gateways. the modular, portable architecture of the analyzer facilitates troubleshooting, analysis, diagnosis and management of multi-vendor, heterogeneous LANs from a single master station. Moreover, as new networking standards emerge, the system can be easily upgraded.
暂无评论