초록 ▼

A method by which a disk-based distributed data storage system is organized for protecting historical records of stored data entities. The method comprises recording distinct states of an entity, corresponding to different moments of time, as separate entity versions coexisting within the distribute

A method by which a disk-based distributed data storage system is organized for protecting historical records of stored data entities. The method comprises recording distinct states of an entity, corresponding to different moments of time, as separate entity versions coexisting within the distributed data storage system, and assigning expiration times to the entity versions independently within each of a plurality of storage sites according to a shared set of rules, before which times deletion is prohibited.

대표청구항 ▼

1. A computer implemented method of by which a distributed disk-based data storage system is organized for protecting historical records of stored data entities, the method comprising the following steps performed by computer hardware: recording distinct states of entities, corresponding to distinct

1. A computer implemented method of by which a distributed disk-based data storage system is organized for protecting historical records of stored data entities, the method comprising the following steps performed by computer hardware: recording distinct states of entities, corresponding to distinct moments of time, as separate entity versions coexisting within the data storage system, the recording comprising:storing blocks of content that constitute the entity versions; and storing references to blocks of stored content to record blocks with identical content;assigning, within the storage system, expiration times to entity versions, before which times both modification and deletion are prohibited;determining a plurality of non-overlapping time intervals and associating with a block of stored content a plurality of reference counts, with each of the plurality of reference counts associated with one of the time intervals;associating with the block of stored content a plurality of expiration times assigned to a plurality of entity versions that the block is a constituent of;recording an approximation of the plurality of expiration times associated with the block of stored content by using the plurality of reference counts associated with the block, with each reference count recording a number of expiration times that fall in the time interval associated with the reference count; anddetermining an expiration time for the block of stored content from the plurality of reference counts, before which time both modification and deletion are prohibited;wherein each of the time intervals has a specified beginning time; andwherein a one of the plurality of reference counts is incremented before its specified beginning time. 2. The method of claim 1 in which incrementing the one of the plurality of reference counts comprises adding a cryptographic hash of a description of a rule to it. 3. The method of claim 1 in which the addition of a new entity version that includes the block of stored content as a constituent causes the one of the plurality of reference counts to be incremented. 4. The method of claim 3 in which the new entity version has an assigned expiration time and an authorized storage client causes the new entity version to be deleted before the assigned expiration time has passed, and after the deletion the expiration time for the block determined from the plurality of reference counts returns to a value it had before the new entity version was added. 5. The method of claim 1 in which the block of stored content is also associated with a reference count that reflects the number of references there are to the block which are not scheduled to expire. 6. The method of claim 1 in which all of the plurality of reference counts are zero and the expiration time determined for the block of stored content depends on the time when the block was created. 7. The method of claim 1 in which the data storage system is distributed and the block of stored content is stored at each of a plurality of storage sites and expiration times are assigned to the plurality of entity versions independently within each of the plurality of storage sites according to a shared set of rules. 8. The method of claim 1 in which an authorized storage client causes a block to be discarded and its space reused when its expiration time has not yet passed. 9. The method of claim 1 in which an authorized storage client overrides the deletion prohibition and causes an entity version to be deleted when its expiration time has not yet passed. 10. The method of claim 9 in which a block of stored content referenced by the deleted entity version is discarded and its storage space is reused. 11. The method of claim 1 in which distinct states of an entity correspond to different moments of time. 12. The method of claim 1 in which each of the plurality of reference counts reflects the number of references there are to the block for entity versions which are scheduled to expire during the associated time interval. 13. The method of claim 1 in which the disk-based data storage system is distributed across a plurality of storage sites and no action taken by a client program that communicates with the distributed disk-based data storage system can cause an expiration time assigned to an entity version, which has not yet passed, to be changed to an earlier time at any correctly operating site of the storage system. 14. The method of claim 13 in which the client program deposits the entity version into the storage system. 15. The method of claim 13 in which the client program causes the expiration time assigned to the entity version to be changed to a later time. 16. The method of claim 13 in which an action taken by the client program causes a value to be assigned to the expiration time. 17. The method of claim 1 in which the disk-based data storage system is distributed across a plurality of storage sites and no action taken by any client program that only communicates with the distributed disk-based data storage system over a wide area network can cause an expiration time assigned to an entity version, which has not yet passed, to be changed to an earlier time at any correctly operating site of the storage system. 18. The method of claim 7 in which expiration times determined for the block of stored content at two different storage sites are different. 19. The method of claim 1 in which the specified beginning time is specified relative to an event. 20. The method of claim 19 in which the event is a creation time associated with the block of stored content. 21. The method of claim 1 in which the specified beginning time is more than a day after the time at which the incrementing occurs. 22. The method of claim 1 in which the block of stored content is an identical portion of the content of each of a plurality of entity versions, and the plurality of reference counts record the number of times the identical portion appears in the plurality of entity versions. 23. The method of claim 1 in which, after the expiration times have been assigned to the plurality of entity versions and before they have passed, a client of the distributed disk-based data storage system is unable to cause the data block to be deleted; and after the expiration times have passed the client is able to cause the data block to be deleted. 24. The method of claim 1 in which an authorized storage client causes one of the plurality of reference counts to be decremented and the expiration time for the block determined from the plurality of reference counts is changed to an earlier time.