초록 ▼

A system and method featuring a collection of model objects which hold data to be shared, and supporting interfaces to read and write the data, to create and destroy collaboration relationships and to attach "view objects" containing application code which will be automatically notified of value cha

A system and method featuring a collection of model objects which hold data to be shared, and supporting interfaces to read and write the data, to create and destroy collaboration relationships and to attach "view objects" containing application code which will be automatically notified of value changes. Each user environment is provided with a model object control entity, referred to herein as a Collaboration Manager, which creates model objects, executes transactions, and creates association objects related to the collaboration. A concurrency control mechanism records the reads and updates of the data and of collaboration relationships within a single transaction site and guarantees that they are applied atomically and consistently to both the local and to all remote replica objects of the collaboration.

대표청구항 ▼

What is claimed is: 1. A computer system for automatically providing consistency of shared computer data in a replica relationshing among a plurality of distributed user sites, each user site being adapted for running at least one application and each site having one of a plurality of collaboration

What is claimed is: 1. A computer system for automatically providing consistency of shared computer data in a replica relationshing among a plurality of distributed user sites, each user site being adapted for running at least one application and each site having one of a plurality of collaboration managers, comprising at each user site: one of said plurality of collaboration managers for creating model objects, for interacting with other of said plurality of collaboration managers from at least one of said plurality of distributed user sites to establish replica relationships with model objects at said at least one user site and for executing transactions that read and update said model objects; a plurality of local model objects created by said collaboration manager for automatically communicating updates to at least one of a plurality of remote model objects at said distributed user sites in replica relationships with said plurilaty of local model objects; interface means for reading and writing said plurality of model objects; a plurality of programmer-specified pessimistic and optimistic view objects for receiving notifications of state changes for said model objects; at least one programming interface for attaching said programmer-specified pessimistic and optimistic view objects to said plurality of local model objects; and wherein said one of said plurality of collaboration managers creates a transaction implementation object having a unique virtual time for a received request; said transaction implementation object invokes said at least one application for executing said request; and said transaction implementation object notifies at least one of said plurality of remote model objects maintaining shared computer information of said executing and, wherein for each of said replica relationships one of said model objects is designated as primary object for replicas, wherein each said primary object is adapted to maintain information regarding notifications received from each said transaction implementation object, and wherein said transaction implementation object notifies said primary object for each of said replica relationships of the version and time for each read and update; each primary object determines whether to issue confirmation of each read and update based upon said maintained information regarding notifications; and said transaction implementation object commits only upon receipt of confirmation from said primary object. 2. The system of claim 1 wherein said plurality of local model objects comprises at least one data model object for responding to requests pertaining to computer information stored at said data model object and at least one association model object for holding data about said replica relationship. 3. The system of claim 1 wherein different distributed user sites are adapted to run different applications utilizing said shared computer information. 4. The system of claim 1 further comprising at least one view proxy object for attaching one for said pessimistic and optimistic view objects to at least one of said plurality of local model objects. 5. The system of claim 1 additionally comprising a plurality of transaction implementation objects, one for each atomic set of operations implemented on said shared computer information. 6. The system of claim 1 further comprising means to immediately notify said optimistic view objects of updates to said local model objects attached thereto. 7. The system of claim 1 further comprising means to immediately notify said pessimistic view objects of committed updates to said local model objects attached thereto. 8. A computer-implemented method for providing automatically updated consistent replicas of shared computer information maintained at a plurality of local and remote model objects at a plurality of distributed user sites based on at least one user transaction request, each site adapted to run at least one application and each site having one of a plurality of collaboration managers for creating model objects, for dynamically establishing replica relationships between model objects at distributed user sites, for interacting with collaboration managers from other of said plurality of user sites, and for executing transactions for reading and writing said model objects, comprising the steps of: one of said plurality of collaboration managers receiving a user request relating to said transaction; said one of said plurality of collaboration managers executing said request and generating an update to one or more of said plurality of local model objects based upon said executing by the steps of: said one of said collaboration managers creating a transaction implementation object having a unique virtual time for said request; said transaction implementation object invoking said at least one application for executing said request; said transaction implementation object notifying at least one of said plurality of remote model objects maintaining said shared computer information of said execution; said one or more of said plurality of local model objects updating said shared computer information at its user site; said one or more of said plurality of local model objects notifying their replicas at remote user sites of said update; each of said remote model objects updating said shared computer information at its user site; and executing decentralized concurrency control to assure atomicity of said transaction and consistency of said replicas, wherein for each replica relationship one of said model objects is designated as primary object for said replicas, wherein said primary objects are adapted to maintain information regarding notifications received from said transaction implementation object, and wherein said concurrency control comprises the steps of: said transaction implementation object notifying said primary object for each replica relationship of the version and time for each read and update; each of said primary objects determining whether to issue confirmation of each read and update based upon said maintained information regarding notifications; and said transaction implementation object committing only upon receipt of confirmation from said primary objects. 9. A computer-implemented method for providing automatically updated consistent replicas of shared computer information maintained at a plurality of local and remote model objects at a plurality of distributed user sites based on at least one user transaction request, each site adapted to run at least one application and each site having one of a plurality of collaboration managers for creating model objects, for dynamically establishing replica relationships between model objects at distributed user sites, for interacting with collaboration managers from other of said plurality of user sites, and for executing transactions for reading and writing said model objects, wherein each of said user sites comprises at least one user interface object comprising one of pessimistic and optimistic view objects attached thereto for providing user notification of updates to said shared computer information and further comprising the step of providing notification of said update to said user at said user interface object, said method comprising the steps of: one of said plurality of collaboration managers receiving a user request relating to said transaction; said one or more of said plurality of local model objects updating said shared computer information at its user site; said one or more of said plurality of local model objects notifying their replicas at remote user sites of said update; each of said remote model objects updating said shared computer information at its user site; and executing decentralized concurrency control to assure atomicity of said transaction and consistency of said replicas, wherein an optimistic view is attached to at least one of said plurality of model objects and wherein said providing notification of said update comprises notifying said optimistic view immediately after each transaction on said at least one of said plurality of model objects completes local execution. 10. A computer-implemented method for proving automatically updated consistent replicas of shared computer information maintained at a plurality of local and remote model objects at a plurality of distributed user sites based on at least one user transaction request, each site having one of a plurality of collaboration managers for creating model objects, for dynamically establishing replica relationships between model objects at distribute user sites, for interacting with collaboration managers from other of said plurality of user sites, and for executing transactions for reading and writing said model objects, wherein each of said user sites of pessimistic and optimistic view objects attached thereto for providing user notification of updates to said shared computer information and further comprising the step of providing notification of said update to said user at said user interface objects, said method comprising the steps of: one of said plurality of collaboration managers receiving a user request relating to said transaction; said one or more of said plurality of local model objects updating said shared computer information at its user site; said one or more of said plurality of local model objects notifying their replicas at remote user sites of said update; each of said remote model objects updating said shared computer information at its user site; and executing decentralized concurrency control to assure atomicity of said transaction and consistency of said replicas, wherein a pessimistic view is attached to at least one of said plurality of model objects and wherein said providing notification of said update comprises notifying said pessimistic view of each committed update to its attached model object.