Method and distributed database file system for implementing self-describing distributed file objects

• US 5,768,532 A
• Filed: 06/17/1996
• Issued: 06/16/1998
• Est. Priority Date: 06/17/1996
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Megerian, Mark Gregory
• Primary Examiner(s): Geckil, Mehmet B.
• Application Number:
  US08/664,706
• Time in Patent Office: 729 Days
• Field of Search: 395/220.31, 395/200.47, 395/133, 395/62, 395/200.75, 711/216, 707/10, 707/103, 707/104, 370/409
• US Class Current: 709/245
• CPC Class Codes: G06F 16/9014 hash tables
• CPC Class Codes: Y10S 707/99945 Object-oriented database structure processing
• CPC Class Codes: Y10S 707/99948 Application of database or data structure, e.g. distributed, multimedia, or image

Abstract

A method and apparatus are provided for implementing self-describing file objects. A node group is created for defining multiple computer systems for storing data. A hash algorithm for applying to data records is identified. A partition distribution map for distributing data to each of the multiple computer systems utilizing a set of predetermined hash algorithm results and remote system information for each of the multiple computer systems are identified. A file object is created in each of the multiple computer systems. Each the file objects includes the hash algorithm, the partition distribution map, and the remote system information. A data record is inserted into one of the distributed file objects by receiving the data record, applying the hash algorithm to the received data record, comparing the hash algorithm result with the partition distribution map to identify the particular computer system for the data record, utilizing the system information to establish connection to that system. The file objects are fully self-describing, eliminating the need for additional objects to be addressed, opened, paged into memory or the like.

Claims

1. A method for implementing self-describing file objects comprising the steps of:

creating a node group to define multiple computer systems for storing the file object;

identifying a hash algorithm for applying to data records;

identifying a partition distribution map for distributing data to each of said multiple computer systems utilizing a set of predetermined hash algorithm results;

identifying remote system information for each of said multiple computer systems; and

creating a file object in each of said multiple computer systems;

each said file object including said hash algorithm, said partition distribution map, and said remote system information.

2. A method for implementing self-describing file objects as recited in claim 1 wherein said step of creating a file object in each of said multiple computer systems;

each said file object including said hash algorithm, said partition distribution map, and said remote system information includes the steps of establishing a connection to each remote computer system and storing said file object in each said remote computer system.

3. A method for implementing self-describing file objects as recited in claim 1 wherein said step of identifying said partition distribution map includes the steps of utilizing a set of possible hash values for said identified hash algorithm and providing an equal distribution of said possible hash values to said multiple computer systems.

4. A method for implementing self-describing file objects as recited in claim 1 wherein said step of identifying said partition distribution map includes the steps of utilizing a set of possible hash values for said identified hash algorithm and providing a user selected distribution of said possible hash values to said multiple computer systems.

5. A method for implementing self-describing file objects as recited in claim 1 further includes the steps of receiving a data record;

applying said hash algorithm to said received data record to generate a hash algorithm result;

comparing said hash algorithm result with said partition distribution map to identify one of said multiple computer systems; and

inserting said received data record into said file object of said identified computer system.

6. A method for implementing self-describing file objects as recited in claim 5 includes the step of establishing a connection to said identified computer system utilizing said remote system information.

7. Apparatus for implementing self-describing file objects comprising:

means for creating a node group to define multiple computer systems for storing the file object;

identifying a hash algorithm for applying to data records;

means for identifying a partition distribution map for distributing data to each of said multiple computer systems utilizing a set of predetermined hash algorithm results;

means for identifying remote system information for each of said multiple computer systems; and

means for creating a file object in each of said multiple computer systems;

each said file object including said hash algorithm, said partition distribution map, and said remote system information.

8. Apparatus for implementing self-describing file objects as recited in claim 7 wherein said means for means for creating said file object in each of said multiple computer systems includes means for establishing a connection to each remote computer system and for storing said file object in each said remote computer system.

9. Apparatus for implementing self-describing file objects as recited in claim 7 further includes means for receiving user selections and wherein said means for identifying said partition distribution map for distributing data to each of said multiple computer systems utilizing said set of predetermined hash algorithm results is responsive to said user selections to define said partition distribution map.

10. Apparatus for implementing self-describing file objects as recited in claim 7 wherein said means for identifying said partition distribution map for distributing data to each of said multiple computer systems utilizing said set of predetermined hash algorithm results includes means for providing an equal distribution of said predetermined hash values to said multiple computer systems.

11. Apparatus for implementing self-describing file objects as recited in claim 7 further includes means for receiving a data record;

means for applying said hash algorithm to said received data record to generate a hash algorithm result;

means for comparing said hash algorithm result with said partition distribution map to identify one of said multiple computer systems.

12. Apparatus for implementing self-describing file objects as recited in claim 11 further includes means for inserting said received data record into said file object of said identified computer system.

13. Apparatus for implementing self-describing file objects as recited in claim 11 further includes means for identifying a remote computer system, means for establishing a connection to said remote computer system and means for inserting said received data record into said file object of said identified remote computer system.

14. A computer program product for use in a multiple computer system, each computer system having a processor, a memory, and a network adapter, the computer program product comprising:

a record medium;

means, recorded on said recording medium, for creating a node group to define multiple computer systems for storing the file object;

means, recorded on said recording medium, for identifying a hash algorithm for applying to data records;

means, recorded on said recording medium, for identifying a partition distribution map for distributing data to each of said multiple computer systems utilizing a set of predetermined hash algorithm results;

means, recorded on said recording medium, for identifying remote system information for each of said multiple computer systems; and

means, recorded on said recording medium, for creating a file object in each of said multiple computer systems;

each said file object including said hash algorithm, said partition distribution map, and said remote system information.

15. A computer program product as recited in claim 14 wherein said means, recorded on said recording medium, for creating said file object in each of said multiple computer systems includes means, recorded on said recording medium, for establishing a connection to each remote computer system and, for storing said file object in each of said multiple computer systems.

16. A computer program product as recited in claim 14 further includes means, recorded on said recording medium, for receiving a data record;

means, recorded on said recording medium, for applying said hash algorithm to said received data record to generate a hash algorithm result;

means, recorded on said recording medium, for comparing said hash algorithm result with said partition distribution map to identify one of said multiple computer systems; and

means, recorded on said recording medium, for inserting said data record into said file object of said identified computer system.