Techniques for synchronizing data store tables

• US 7,739,234 B1
• Filed: 12/15/2005
• Issued: 06/15/2010
• Est. Priority Date: 12/15/2005
• Status: Active Grant

• Current Assignee: Teradata US, Inc. (Teradata Corporation)
• Original Assignee: Teradata US, Inc. (Teradata Corporation)
• Inventors: Pritchard, Sean, Canna, Jeffrey, Peres-da-Silva, Anil, Chapman, Alex, Tackett, Dennis, Shah, Anish, Chaves, Michael, Lee, Harold, Figge, Jason
• Primary Examiner(s): Cottingham; John R.
• Assistant Examiner(s): Uddin; Mohammed R
• Application Number:
  US11/304,173
• Time in Patent Office: 1,643 Days
• Field of Search: 707/4, 707/203
• US Class Current: 707/610
• CPC Class Codes: G06F 16/902 using more than one table in sequence, i.e. systems with three or more layers

Abstract

Techniques are presented for synchronizing data store tables. Elements of first tables are linked together to form dependent chains. The dependencies are not housed with the elements in the first tables; rather, the dependencies are housed in multiple second tables. Each second table maps one of the elements from one of the first tables to another of the elements in another of the first tables.

Claims

1. A method implemented in a computer-readable medium to execute on a computer, the method, comprising:

receiving, by the computer, an update to an entry in a first original table of a relational database, wherein the entry is a particular field of a particular row of the first original table;

creating, by the computer, a new entry within the first original table to house the updated entry;

determining, by the computer, the entry is dependent on a chain of linked other original tables, the chain of linked other original tables forms a linked list, which includes pointers from each other table entry to another entry in another of the other original tables, the chain is also represented via another separate and independent set of mapping tables for transitions that maintain dependencies independent from the first table and the other original tables that house the entry and other entries, the other original tables do not include linked dependencies among one another, the dependencies are housed in the mapping tables and just two tables are accessed to update and to reflect the update to the entry occurring in the first original table, these two updates are a first update to the first original table to show the update for the entry and a second update to a next to last mapping table in the mapping tables, the next to last mapping table is a particular mapping table that points to the entry being updated in the first original table;

traversing, by the computer, the chain to the next to last mapping table that points to the entry in the first original table;

acquiring, by the computer, the particular mapping table for the next to last mapping table; and

adding, by the computer, an updated entry in the particular mapping table to link to the updated entry appearing in the first original table.

2. The method of claim 1, wherein creating further includes incrementing a versioning number associated with the entry and associating the incremented versioning number with the new entry.

3. The method of claim 1, wherein determining further includes identifying the chain as an object network, wherein each element of the network is a row in one of the other original tables.

4. The method of claim 3, wherein traversing further includes accessing a root node representing a first element of the object network to begin traversing the object network.

5. The method of claim 1 further comprising, removing, by the computer, the entry from the first original table if the entry is free of dependencies.

6. A method implemented in a computer-readable medium to execute on a computer, the method, comprising:

receiving, by the computer, a reference to a chain of linked elements of original database tables, the chain is a linked list that includes pointers from a root database table to a last database table;

traversing, by the computer, the chain and building a plurality of separate and independent mapping tables, a total number of the mapping tables equal to a total number of the original database tables minus one, and there being one particular mapping table for each pair of database tables of the chain, the chain is associated with each mapping table and represents transitions and each mapping table maintaining dependencies independent from original database tables that house information; and

adding, by the computer, an entry to a particular mapping table that links a first element of that particular mapping table to a second element of a next original database table associated with the chain, the first element included in a particular original database table that precedes the next original database table within the chain and the dependency that links the first element to the second element is defined in the particular mapping table, when an update is made to a field in a particular original database table, the particular original database table and remaining ones of the original database tables do not include dependencies that link the particular original database table and the remaining original database tables rather the dependencies of the linking between the particular original database table and the remaining original database tables are maintained in one of the mapping tables for the update, and when the update to the particular original database table occurs, two updates occur one to the particular original database table and another update to the one mapping table having the dependencies complete the update.

7. The method of claim 6 further comprising, maintaining, by the computer, a versioning number with each element of the chain.

8. The method of claim 6 further comprising, updating, by the computer, the chain when a particular element is modified by updating an appropriate one of the mapping tables.

9. The method of claim 6, wherein receiving further includes receiving the reference as a pointer to a first element of the chain.

10. The method of claim 6 further comprising, including, by the computer, a chain identifier with each of the entries of the mapping tables thereby allowing the mapping tables to be used for multiple chains based on a specific table.

11. The method of claim 6 further comprising, identifying, by the computer, each mapping table as an association between two of the original database tables.