Patent ReferencesThrashing reduction in demand accessing of a data base through an LRU paging buffer pool Method for dynamically allocating LRU/MRU managed memory among concurrent sequential processes Method of inter-process communication in a distributed data processing system Communication for version management in a distributed information service System and method for using cached data at a local node after re-opening a file at a remote node in a distributed networking environment Computer communication system having supplemental formats Single system image uniquely defining an environment for each user in a data processing system Distributed data management mechanism System for accessing remote heterogeneous database including formatting retrieved data into applications program format Data access system for a file access processor InventorApplicationNo. 343477 filed on 11/28/1994US Classes:707/8, Concurrency (e.g., lock management in shared database)710/56, Buffer space allocation or deallocation711/137, Look-ahead711/141CoherencyExaminersPrimary: Kim, Kenneth S.Attorney, Agent or FirmInternational ClassG06F 015/163AbstractNetwork Distributed Caches ("NDCs") (50) permit accessing a named dataset stored at an NDC server terminator site (22) in response to a request submitted to an NDC client terminator site (24) by a client workstation (42). In accessing the dataset, the NDCs (50) form a NDC data conduit (62) that provides an active virtual circuit ("AVC") from the NDC client site (24) through intermediate NDC sites (26B, 26A) to the NDC server site (22). Through the AVC provided by the conduit (62), the NDC sites (22, 26A and 26B) project an image of the requested portion of the named dataset into the NDC client site (24) where it may be either read or written by the workstation 42. The NDCs (50) maintain absolute consistency between the source dataset and its projections at all NDC client terminator sites (24, 204B and 206) at which client workstations access the dataset. Channels (116) in each NDC (50) accumulate profiling data from the requests to access the dataset for which they have been claimed. The NDCs (50) use the accumulated profile data stored in channels (116) to anticipate future requests to access datasets, and, whenever possible, prevent any delay to client workstations in accessing data by asynchronously pre-fetching the data in advance of receiving a request from a client workstation. | |