| 摘要 |
The present invention, generally speaking, provides a method of synchronizing information between a plurality of sites and a central location that solves number of potentially "thorny" problems of distributed database systems in an elegant and efficient manner. The first issue involves sychronizing different copies of an item on machines that are not continuously linked at which each copy may be changed independently. Rather than attempting to formulate a set of synchronization policies generally applicable to all items in all instances, synchronization policies are moved from the database engine to the items themselves. The second issue involves lifetime-knowing, given a complex set of interrelationship of items, when an item may be safely deleted. This problem is addressed by providing an interested party mechanism whereby items may "express interest" in each other. When no item is interested in a particular item, that item may safely "go away".
|
