Showing papers on "Serialization published in 1985"

Synchronization and recovery of actions

Abstract: We introduce an approach to robust computation in distributed systems. This approach is the foundation for reliability in the Clouds decentralized operating system. It is based on atomic actions operating on instances of abstract data types (objects). We present an event-based model of computation in which scheduling of responses to operation invocations is controlled by objects. We discuss an integrated strategy for synchronization and recovery which uses relationships between the abstract states of objects to track dependencies between actions. Serializability is defined in terms of the semantics of operations. This permits high concurrency to be obtained in non-serializable implementations without deviation from serializable abstract behavior. We define a class of schedulers that allows objects to make autonomous scheduling decisions. We present the use of non-serializable operation semantics. Finally, we discuss implementation of the model, including action synchronization, object operation ordering using action-based counting semaphores, and action recovery.

Approaching Distributed Database Implementations Through Functional Programming Concepts

Abstract: The application of functional programming concepts to the data representation and query!ng aspects of databases has been discussed by Shipman and Buneman, et al. respectively. We argue the suitability of a function-based approach to additional aspects of database systems, including updating, transaction serialization, and physical distribution and communication. It is shown how the NmergeH extension of a purely functional model permits seriallzable concurrent "primary site" distribution control. We also present preliminary experimental results which indicate that a reasonable degree of concurrency is attainable from the functional approach.

A Simple Characterization of Database Serializability

Abstract: An interleaved execution of transactions in a database system is serializable if the effect of the execution is equivalent to that of some serial execution of the transactions. We give a simple and intuitive characterization of serializability that brings out the inherent problem of serialization explicitly. We also give another characterization which extends naturally to multiversion serializability.

Synthesis of parallel programs invariants

Abstract: Most current methods for parallel programs design and verification are based on the concept of invariant. However, invariant synthesis is the most difficult part of those methods. This work presents a technique for invariant design usable for both parallel program synthesis and verification. This technique drastically reduces the risks of errors due to a bad statement serialization or to inadequate interprocess synchronization.

File placement and process assignment due to resource sharing in a distributed system

Abstract: This paper presents a number of solutions to the problems of data access, concurrency control, transaction serialization, and deadlock control based on the examples of existing file systems. Also, policies for file placement and process assignment are discussed from the viewpoint of their impact on system performance and reliability. The impact of these policies depends on the solutions to the problems referred above. A simulation model of a file system is introduced. This model is validated using data collected in a small business installation. A number of simulation experiments are presented. It is shown how the decision about where to place a file or to assign a process can be based on the workload's and the system's characterization. An algorithm for improving performance and reliability by file placement and/or process assignment is finally presented.

On the modeling of read and write locks on a distributed system

Abstract: This paper presents a new approach to estimating serialization delays in computer systems. An analytic model is introduced to describe synchronization of accesses to shared resources. The model parameters are estimated from the values of measurable quantities. The probabilities that a shared resource is locked, thereby causing other processes to wait for the resource to be accessible, are calculated, under certain assumptions, as functions of the probabilities of access and of the number of active transactions. The service times of lock servers are calculated as functions of the loads on all service centers. Performance measures applicable to the evaluation of computer systems are introduced and evaluated. An example involving a distributed file system and measurement data collected in a small business installation is given to compare performance measures provided by the simulation and analytic models.

Insertion System and digital data restitution using a time code binds a program recorded on VCR

Abstract: System and insertion of digital data restitution using a time code associated with an audiovisual program recorded on VCR. This system essentially includes nine functional blocks: a main memory 120, an address generator 124, two digital conversion circuits 126, 128, a serialization code circuit 130, a switching circuit 132, a circuit deserializing the videographic code 134, a serial interface circuit 136 and a control circuit 140. a bus 114 connects the blocks. Application television. (CF DRAWING IN BOPI)