Field level replication method

• US 6,138,124 A
• Filed: 06/09/1998
• Issued: 10/24/2000
• Est. Priority Date: 01/11/1996
• Status: Expired due to Term

• Current Assignee: International Business Machines SA (International Business Machines Corporation)
• Original Assignee: International Business Machines SA (International Business Machines Corporation)
• Inventors: Beckhardt, Steven R.
• Primary Examiner(s): Kulik, Paul V.
• Assistant Examiner(s): Corrielus, Jean M.
• Application Number:
  US09/094,216
• Time in Patent Office: 868 Days
• Field of Search: 707/201, 707/8, 711/141, 711/142, 711/146, 395/182.18, 395/670, 395/680
• US Class Current: 1/1
• CPC Class Codes: G06F 16/27 Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
• CPC Class Codes: G06F 16/278 Data partitioning, e.g. horizontal or vertical partitioning
• CPC Class Codes: Y10S 707/99938 Concurrency, e.g. lock management in shared database
• CPC Class Codes: Y10S 707/99952 Coherency, e.g. same view to multiple users

Abstract

In a distributed computing system in which replicas of a document are separately stored and revised, the document containing data arranged in a number of fields, a method for replicating data contained in a revised document replica to the other of the replicas by replicating only the field or fields which have been revised since an earlier replication. The method includes the steps of dynamically maintaining a two byte document sequence number for each of the document replicas representing the number of revisions made to the replicas, and dynamically maintaining a one byte field sequence number for each of the fields in the replicas. The field sequence numbers for revised fields are set equal to the lower byte of the current document sequence number. When a replication is desired, the method involves determining which of the document replicas is more recently revised, determining which of the fields in the more recently revised replica are more recently revised than the corresponding fields in the other replica based on some or all of the field and document sequence numbers, and replicating only the data contained in each more recently revised fields to the corresponding fields in the other replica. The system can handle document sequence numbers greater than 256, in which field sequence numbers can no longer be accurately represented by one byte, by artificially incrementing field sequence numbers for unrevised fields and interpreting all field sequence numbers as having a value within 256 of the document sequence number.

Claims

1. In a distributed computing environment in which each of two documents contains data arranged in a plurality of fields, at least some of the fields of one of the documents corresponding to at least some of the fields of the other of the documents, and at least two fields in each document being revisable, a method for replicating data contained in one of the documents to the other of the documents by replicating only the field or fields which have been revised, the method comprising:

dynamically maintaining a document sequence number for each of the documents representing the number of revisions made to each of the documents;

dynamically maintaining a field sequence number for each of the revisable fields;

when a replication is to be made, determining which of the documents is more recently revised;

determining which one or more fields in the more recently revised document are more recently revised than the corresponding one or more fields in the other of the documents based on some or all of the field sequence numbers and document sequence numbers; and

replicating the data contained in each more recently revised field to the corresponding field in the other of the documents.

2. The method of claim 1 wherein the step of determining which of the documents is more recently revised comprises comparing the document sequence numbers of the documents to determine which is greater.

3. The method of claim 1 further comprising stamping each document with a revision time stamp and updating the revision time stamp when each document is revised, and wherein the step of determining which of the documents is more recently revised comprises comparing the revision time stamps of the documents to determine which represents a later time.

4. In a distributed computing system in which replicas of a document are separately stored and revised, the document containing data arranged in a plurality of fields and being periodically replicated, a method for replicating a document comprising:

dynamically maintaining a document sequence number for each of the document replicas representing the number of revisions made to each replica;

dynamically maintaining a field sequence number for each of the fields in each replica;

determining which of the document replicas is more recently revised;

determining which field or fields in the more recently revised replica have been revised since a previous replication of the document based on some or all of the field and document sequence numbers; and

replicating the data contained in each revised field to the corresponding field in the other replica.

5. A method comprising:

storing a document sequence number for a document and a field sequence number for each of a plurality of revisable fields contained within the document;

incrementing the document sequence number each time a revision to one or more of the fields in the document is registered; and

setting a first new value for each field sequence number corresponding to each of one or more fields in the document which have been revised, wherein the new value is set based on the incremented document sequence number.

6. A method for replicating data from a first document to a second document, each document containing data arranged in a plurality of fields, at least some of the fields of the first document corresponding to at least some of the fields of the second document, the first document having a first document sequence number representing the number of revisions to the first document, the second document having a second document sequence number representing the number of revisions to the second document, the plurality of fields each having a field sequence number, the method comprising the steps of:

identifying one or more field sequence numbers of the plurality of field sequence numbers of the first document having a higher value than the second document sequence number; and

replicating the data in each field of the first document having a field sequence number of higher value than the second document sequence number to the corresponding field of the second document.