ITEM-SET KNOWLEDGE FOR PARTIAL REPLICA SYNCHRONIZATION

• US 20080294701A1
• Filed: 05/21/2007
• Published: 11/27/2008
• Est. Priority Date: 05/21/2007
• Status: Abandoned Application

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Rodeheffer, Thomas L., Wobber, Edward P., Terry, Douglas, Saraswati, Venugopalan Ramasubramanian
• Application Number:
  US11/751,478
• Publication Number:
  US 20080294701A1
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06F 16/273 Asynchronous replication or reconciliation
• CPC Class Codes: H04L 67/1095 Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes

Abstract

A system is disclosed for synchronizing partially-replicated collections while keeping synchronization overhead low by using the concept of item-set knowledge. Item-set knowledge uses knowledge fragments, which associate knowledge vectors with item-sets. An item-set consists of an explicitly represented list of items. In a partial replica, this item set may be the items known to a replica within which a filter is applied limiting the items known to some subset of the overall items in the collection.

Claims

1. A computer implemented method of synchronizing a plurality of replicas having a collection of items, the method comprising the steps of:

(a) receiving a sync request from a first replica at a second replica, said request containing a plurality of item-set knowledge fragments, said plurality of item-set knowledge fragments indicating versions of items of which the first replica is aware, one of said fragments referring to a plurality of items;

(b) transmitting from the second replica to the first replica versions of items which are not known to the first replica; and

(c) transmitting from the second replica to the first replica a plurality of item-set knowledge fragments, said plurality of item-set knowledge fragments indicating versions of items learned by the first replica, one of said fragments referring to a plurality of items.

2. A computer implemented method as recited in claim 1, further comprising defragmenting the result of the item-set knowledge fragments known to the first replica before said step (c) and the knowledge fragments received by the first replica in said step (c) to obtain knowledge fragments that are reduced in size or number.

3. A computer implemented method as recited in claim 2, wherein said step of defragmenting comprises the step of replacing a knowledge fragment in the first replica that existed prior to said step (c) with a knowledge fragment received in said step (c), where the knowledge fragment received in said step (c) dominates the knowledge fragment existing before said step (c).

4. A computer implemented method as recited in claim 2, wherein said step of defragmenting comprises the step of discarding a knowledge fragment received in said step (c) from the second replica where a knowledge fragment existing in the first replica prior to said step (c) dominates the knowledge fragment received in said step (c).

5. A computer implemented method as recited in claim 1, wherein a replica maintains a set of knowledge fragments for versions of items that it stores.

6. A computer implemented method as recited in claim 1, wherein a replica maintains a set of knowledge fragments for versions of items that it does not store.

7. A computer implemented method of synchronizing a plurality of replicas having a collection of items, the plurality including a first replica having a first set of items, and a second replica having a second set of items, the method comprising the steps of:

(a) associating a knowledge vector with the first set of items in the first replica to define a first knowledge fragment;

(b) associating a knowledge vector with the second set of items in the second replica to define multiple knowledge fragments;

(c) receiving a sync request from the first replica at the second replica;

(d) transmitting from the second replica to the first replica versions of items which are not known to the first replica; and

(e) transmitting from the second replica to the first replica the multiple knowledge fragments for addition to the first knowledge fragment.

8. A computer implemented method as recited in claim 7, wherein the second replica is a partial replica and the second set of items is a smaller subset of items than the general set of items in the collection.

9. A computer implemented method as recited in claim 7, wherein the first replica is a partial replica and the first set of items is a smaller subset of items than the general set of items in the collection.

10. A computer implemented method as recited in claim 9, wherein said step (d) of transmitting from the second replica to the first replica versions of items which are not known to the first replica comprises the step of transmitting versions of items that are not known to the first replica and which satisfy a filter applied within the second replica.

11. A computer implemented method as recited in claim 7, wherein said step (e) of transmitting from the second replica to the first replica the multiple knowledge fragments for addition to the first knowledge fragment comprises the step of defragmenting the additive result of the first knowledge fragment and the multiple knowledge fragments.

12. A computer implemented method of synchronizing between a collection of replicas having a set of items, the collection including a first replica having a first set of items, and a second replica having a second set of items, the method comprising the steps of:

(a) associating a knowledge vector, K₁, with the first set of items, S₁, in the first replica to define a first knowledge fragment, S₁;

K₁;

(b) associating a knowledge vector, K₂, with the second set of items, S₂, in the second replica to define a second knowledge fragment, S₂;

K₂;

(c) receiving a sync request from the first replica at the second replica;

(d) transmitting from the second replica to the first replica the second knowledge fragment for addition to the first knowledge fragment; and

(e) defragmenting the additive result of the first and second knowledge fragments obtained in said step (d).

13. A computer implemented method as recited in claim 12, wherein at least one of the first and second replicas are partial replicas.

14. A computer implemented method as recited in claim 12, wherein said step (e) of defragmenting the additive result of the first and second knowledge fragments comprises the step of discarding the S₁:

K₁ knowledge fragment where K₁ is a subset of K₂, and S₁ is a subset of S₂ resulting in a knowledge fragment of S₂;

K₂.

15. A computer implemented method as recited in claim 12, wherein said step (e) of defragmenting the additive result of the first and second knowledge fragments comprises the step of discarding the S₁:

K₁ knowledge fragment where K₁ is a subset of K₂, and S₁ equals or is a subset of S₂ resulting in a knowledge fragment of S₂;

K₂.

16. A computer implemented method as recited in claim 12, wherein said step (e) of defragmenting the additive result of the first and second knowledge fragments comprises the step of discarding the S₁:

K₁ knowledge fragment where K₁ equals K₂, and S₁ is a subset of S₂ resulting in a knowledge fragment of S₂;

K₂.

17. A computer implemented method as recited in claim 12, wherein said step (e) of defragmenting the additive result of the first and second knowledge fragments comprises the step of discarding the S₂:

K₂ knowledge fragment where K₁ equals K₂, and S₂ is equal to or a subset of S₁ resulting in a knowledge fragment of S₁;

K₁.

18. A computer implemented method as recited in claim 12, wherein said step (e) of defragmenting the additive result of the first and second knowledge fragments comprises the step of keeping S₁:

K₁+S₂;

K₂ as the resulting knowledge fragment where K₁ is not equal to K₂, and S₁ is not equal to S₁.

19. A computer implemented method as recited in claim 12, wherein the second replica is a partial replica and the second set of items is a smaller subset of items than the general set of items in the collection.

20. A computer implemented method as recited in claim 12, wherein the first replica is a partial replica and the first set of items is a smaller subset of items than the general set of items in the collection.