Location management is an important problem in distributed mobile computing. Location management consists of location updates, searches and search-updates. An update occurs when a mobile host changes location. A searc...
详细信息
Location management is an important problem in distributed mobile computing. Location management consists of location updates, searches and search-updates. An update occurs when a mobile host changes location. A search occurs when a mobile host needs to be located. A search-update occurs after a successful search. Various strategies can be designed for search, update and search-update. Static location management uses one combination of search, update and search-update strategies throughout the execution. Simulations were carried out to evaluate the performance of different static strategies for various communication and mobility patterns. Simulation results indicate that performing search-updates significantly reduces the message overhead of location management.
The integration of mobile/portable computing devices within existing data networks can be expected to spawn distributed applications that execute on mobile hosts (MHs). For reliability, it is vital that the global sta...
详细信息
The integration of mobile/portable computing devices within existing data networks can be expected to spawn distributed applications that execute on mobile hosts (MHs). For reliability, it is vital that the global state of such applications be checkpointed from time to time. A global checkpoint consists of a set of local checkpoints, one per participant. This paper first identifies the problems in recording a consistent global state of mobile distributed applications. The location of a MH within the static network varies with time and therefore, a MH will first need to be located ('searched') in order to obtain its local checkpoint. Moreover, MHs often (voluntarily) disconnect from the network;a disconnected MH is not reachable from the rest of the network. This means that a (disconnected) MH may not be available to provide its local checkpoint. Lastly, a MH is not equipped with stable storage;disk space at a MH is not considered stable due to vulnerability of MHs to loss, theft and physical damage. Therefore, an alternative stable repository is required to save local checkpoints of MHs. This paper presents a checkpointing algorithm for MHs that satisfies these constraints.
We present a scalable garbage collection scheme for systems that store objects at multiple servers while clients run transactions on locally cached copies of objects. It is the first scheme that provides fault toleran...
详细信息
We present a scalable garbage collection scheme for systems that store objects at multiple servers while clients run transactions on locally cached copies of objects. It is the first scheme that provides fault tolerance for such a system: Servers recover from failures and retrieve information needed for safe garbage collection;clients do not recover from failures, yet the scheme is able to reclaim objects referenced only from failed clients. The scheme is optimized to reduce overhead on common client operations, and it provides fault tolerance by doing work in the background and during client operations that are infrequent.
Many organisations are now planning to move their operations from total reliance on centralised databases towards distributed environments which may involve the interoperability of a number of heterogeneous databases....
详细信息
Many organisations are now planning to move their operations from total reliance on centralised databases towards distributed environments which may involve the interoperability of a number of heterogeneous databases. This study looks at a particular case for an international financial institution, with the likely performance of a proposed globally distributed federated database being compared with the performance of the current centralised system. The performance model developed includes submodels for transaction structure and management, user workload, and distributed heterogeneous databases. Simulations focus on response times for a particular class of credit control/deal entry transactions in the presence of a background load. The results demonstrate that the proposed federated database outperforms the current centralised system, and that this is achievable using currently available technology.
Many applications need transaction-consistent pictures of past states of the database. These applications use the commit time of the transaction to timestamp the data. When a transaction is distributed, the cohorts mu...
详细信息
Many applications need transaction-consistent pictures of past states of the database. These applications use the commit time of the transaction to timestamp the data. When a transaction is distributed, the cohorts must vote on a commit time and the coordinator must choose a commit time based on the votes of the cohorts. This implies that timestamps are applied after commit. Until the timestamps are on all the records, one must keep a table of all the committed transaction identifiers and their commit times. The main problem solved is that of determining when all timestamps corresponding to a given transaction have been placed in the records so that a committed transaction entry can be erased from the table. This information must be stable. Logging and recovery details are included.
We propose a new network architecture for personal communication services that provides transparent connectivity to different physical networks and user service customization. User information in this network is manag...
详细信息
We propose a new network architecture for personal communication services that provides transparent connectivity to different physical networks and user service customization. User information in this network is managed independently by distributed personal agents(PAs) which are identified by their user's personal ID. PAs interact with each other to establish the best communication paths between their users according to their states.
For a data sharing database system, substantial cost is incurred for global locking, to support both transactions and cache management. Replacing global locks with local locks managed by local lock managers offers the...
详细信息
For a data sharing database system, substantial cost is incurred for global locking, to support both transactions and cache management. Replacing global locks with local locks managed by local lock managers offers the opportunity to substantially reduce locking overhead. To do this requires the exploitation of global covering locks. We discuss covering in general, and describe the conflicts required for both covering and intention locks. We then describe how to generate new covering and intention modes when logical and physical resources are equated in a data sharing system, hence reducing locking overhead. The new intention modes, because they conflict, permit cache management without losing fine grained concurrency. Fine grained concurrency with combined resources was not previously possible.
Various existing dynamic channel assignment (DCA) strategies are introduced first and a new distributed DCA algorithm is proposed in the paper. Based on local observations of the reuse cluster, the algorithm attempts ...
详细信息
Various existing dynamic channel assignment (DCA) strategies are introduced first and a new distributed DCA algorithm is proposed in the paper. Based on local observations of the reuse cluster, the algorithm attempts to minimize the reuse distance of the allocated channels and to keep channel reassignments under control. System performance of the proposed algorithm is analyzed and compared with other DCA algorithms. Simulation results show that, under certain conditions, the algorithm may approximate blocking performance of the centrally optimized Maximum Packing (MP) algorithm.
Many new database applications require very large volumes of data. Mariposa is a data base system under construction at Berkeley responding to this need. Mariposa objects can be stored over thousands of autonomous sit...
详细信息
Many new database applications require very large volumes of data. Mariposa is a data base system under construction at Berkeley responding to this need. Mariposa objects can be stored over thousands of autonomous sites and on memory hierarchies with very large capacity. This scale of the system leads to complex query execution and storage management issues, unsolvable in practice with traditional techniques. We propose an economic paradigm as the solution. A query receives a budget which it spends to obtain the answers. Each site attempts to maximize income by buying and selling storage objects, and processing queries for locally stored objects. We present the protocols which underlie the Mariposa economy.
This paper describes a distributed radio port controller architecture to provide PCS services in a micro-cellular environment. The architecture allows for rapid introduction of PCS services for mobile users using exis...
详细信息
This paper describes a distributed radio port controller architecture to provide PCS services in a micro-cellular environment. The architecture allows for rapid introduction of PCS services for mobile users using existing central office switches. Circuit and packet data services are efficiently integrated with voice services. We present the response of the vendor community to this architecture, and discuss the feasibility of its deployment. Integration of this architecture with cellular systems is explored to provide national coverage. Issues regarding seamless handovers between pedestrian and vehicular speeds in a two-tier radio system are discussed. Finally, the preliminary result of testing to validate the architecture by a simulator in a testbed is presented.
暂无评论