System and method for peer-to-peer grid based autonomic and probabilistic on-demand backup and restore

• US 7,584,226 B2
• Filed: 05/24/2005
• Issued: 09/01/2009
• Est. Priority Date: 05/24/2005
• Status: Expired due to Fees

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Goldstein, Colin, Fatula, Joseph J. Jr.
• Primary Examiner(s): Cottingham; John R.
• Assistant Examiner(s): Uddin; Mohammed R
• Application Number:
  US11/135,876
• Publication Number:
  US 20060271601A1
• Time in Patent Office: 1,561 Days
• Field of Search: 707/204, 707/202
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1451 by selection of backup contents
• CPC Class Codes: G06F 11/1464 for networked environments
• CPC Class Codes: G06F 11/1469 Backup restoration techniques
• CPC Class Codes: Y10S 707/99953 Recoverability
• CPC Class Codes: Y10S 707/99955 Archiving or backup

Abstract

A system and method for peer-to-peer grid based autonomic and probabilistic on-demand trackless backup and restore are disclosed. The on-demand trackless backup-restore system includes a plurality of nodes connected to a network communications channel. The plurality of nodes in the on-demand trackless backup-restore system includes at least one node that is a source client configured to initiate a trackless data backup-restore request and includes one or more nodes that are target clients configured to perform the trackless data backup-restore request. Further, the source client in the on-demand trackless backup-restore system is configured to establish direct connections with one or more of the target clients that are available to perform the trackless data backup-restore request.

Claims

1. An on-demand data backup-restore system for backing up and restoring data in a grid computing system, said system comprising:

a grid computing system network communications channel; and

a plurality of client nodes connected to said grid computing system network communications channel for said grid computing system, said plurality of client nodes having at least one central processing unit for executing at least one grid application installed thereon, and for executing a trackless data backup-restore software installed on each of said plurality of client nodes for performing an on-demand data backup-restore request, said trackless data backup-restore software providing a respective trackless unique file identifier name to each of one or more data files selected for said on-demand data backup-restore request, at least one of said plurality of client nodes being a source client configured to initiate said on-demand data backup-restore request for either backing up data for said grid application on said grid computing system or for retrieving data stored for said grid application on said grid computing system, one or more of said plurality of client nodes being target clients configured to perform said on-demand data backup-restore request for said grid application on said grid computing system and at least one of said plurality of client nodes being a super node configured to broker an initial communication for said on-demand backup-restore request for said grid application between said source client and said target clients on said grid computing system for backing-up data for said grid application or restoring data for said grid application on said grid computing system, wherein said source client establishes one or more simultaneous sessions with a plurality of available target clients among said target clients, wherein said source client provides to said plurality of available target clients a trackless unique file identifier name corresponding to each of one or more data files selected for said on-demand data backup-restore request, wherein said source client sends one or more data file fragments for a respective data file selected to said plurality of available target clients, and wherein said plurality of available target clients perform said on-demand data backup-restore request using said trackless unique file identifier name provided for said respective data file selected corresponding to each of said one or more data file fragments, and wherein said source client does not track a respective location of said one or more data file fragments backed up on or retrieved from said plurality of available target clients.

2. The system according to claim 1, further comprising:

a subscription manager configured to manage a contractual subscription for each of said plurality of client nodes either backing up data for said grid application on said grid computing system or retrieving data stored for said grid application on said grid computing system.

3. The system according to claim 1, further comprising:

a super node list provider for supplying a super node list that includes said at least one super node configured to broker said initial communication for said on-demand backup-restore request between said source client and said target clients on said grid computing system.

4. The system according to claim 3, wherein said super node on said grid computing system further comprises:

a communication component for communicating said on-demand data backup-restore request received from said source client to said target clients on said grid computing system, wherein said communication component includes a data backup request module for communicating an on-demand data backup request from said source client to said target clients for backing up data, a data retrieval request module for communicating an on-demand data retrieval request from said source client to said target clients for retrieving data, a data restore request module for communicating an on-demand data restore request form said source client to said target clients for restoring data and a list generation module for generating a list of said available target clients among said target clients that are available for performing said on-demand data backup-restore request.

5. The system according to claim 1, wherein each of said plurality of client nodes on said grid computing system further comprises:

a global client profile for storing one or more global parameters for said grid computing system, a local client profile for storing one or more local parameters for a client node of said plurality of client nodes, a data assembly record for storing information pertaining to assembly of a single backup file stored on said grid computing system and a log for storing a backup-restore history of operations performed on said grid computing system over a period of time.

6. The system according to claim 1, wherein each of said plurality of client nodes further comprises:

an on-demand data backup-restore component that includes a communication module for establishing direct connections with said plurality of available target clients for backing up data or for retrieving backup data, a profile management module for managing default and unique client profiles that are specific to each of said plurality of client nodes, a data storage module for storing backup data received from said plurality of client nodes, a data retrieval module for retrieving backup data from said plurality of client nodes, a data restoration module for restoring data in an original data file backed up on said plurality of client nodes, a data assembly module for assembling data into a format similar to said original data file backed up on said plurality of client nodes, a data compression module for compressing data to be backed up on said plurality of client nodes, a data encryption module for encrypting data to be backed up on said plurality of client nodes, a data redundancy module for duplicating data to be backed up on said plurality of client nodes and a unique file identifier generation module for generating a respective trackless unique file identifier name corresponding to each original data file backed up on or retrieved from said plurality of client nodes on said grid computing system.

7. The system according to claim 1, wherein said on-demand data backup-restore request comprises one of:

backing up a data file, backing up one or more of a plurality of data file fragments, restoring a data file, retrieving a data file, retrieving one or more of a plurality of data file fragments, restoring one or more of a plurality of data file fragments and restoring a client node of said plurality of client nodes that has experienced a loss of data.

8. A method for trackless backup and restoration of data on-demand across a grid computing system cluster of client nodes, said method comprising the steps of:

providing said grid computing system cluster of client nodes connected to a grid computing system network communications channel, said grid computing system cluster of client nodes having a grid computing software installed thereon for executing at least one grid application, said grid computing system cluster of client nodes having installed thereon a trackless data backup-restore software for backing up and restoring data, said grid computing system cluster of client nodes including a source client configured to initiate on-demand a trackless data backup-restore request for one or more data file fragments from one or more data files and a plurality of target clients configured to perform on-demand said trackless data backup-restore request for said one or more data file fragments, wherein each of said one or more data file fragments has a respective trackless unique file identifier name;

seeking a plurality of grid computing system available target clients from said plurality of target clients that are available for performing said trackless data backup-restore request for said one or more data file fragments for said grid application;

establishing direct connections with a subset of said plurality of grid computing system available target clients that are available for performing said trackless data backup-restore request for said one or more data file fragments corresponding to respective trackless unique file identifier names; and

directing said trackless data backup-restore request by employing said subset of said plurality of grid computing system available target clients for performing said trackless data backup-restore request for said one or more data file fragments for said grid application;

wherein said subset of said plurality of grid computing system available target clients perform said trackless data backup-restore request using said respective trackless unique file identifier name provided corresponding to said one or more data file fragments of a respective data file, and wherein said source client does not track a respective location of said one or more data file fragments backed up on or retrieved from said plurality of grid computing system available target clients.

9. A method according to claim 8, wherein said directing step further comprises the steps of:

transmitting said trackless data backup-restore request for said one or more data file fragments having said respective trackless unique file identifier names to said subset of said plurality of grid computing system available target clients on said grid computing system; and

receiving a response to said trackless data backup-restore request from said subset of said plurality of grid computing system available target clients that have said one or more data file fragments corresponding to said respective trackless unique file identifier names transmitted, wherein said source client can determine if sufficient data was backed up or retrieved successfully.

10. A method according to claim 8, wherein said trackless data backup-restore request comprises one of:

backing up a data file, backing up one or more of a plurality of data file fragments, restoring a data file, retrieving a data file, retrieving one or more of a plurality of data file fragments, restoring one or more of a plurality of data file fragments and restoring a client node of said cluster of client nodes that has experienced a loss of data.

11. A method according to claim 10, wherein said trackless data backup-restore request is for backing up said one or more data files, and said directing step further comprises the steps of:

selecting said one or more data files to be backed up by said subset of said grid computing system available target clients on said grid computing system;

computing a trackless unique file identifier name for said each of said one or more data files selected;

adding redundancy to said one or more data files selected based on a plurality of profile settings;

fragmenting into a plurality of data file fragments each of said one or more data files selected to which redundancy has been added;

encrypting each of said plurality of data file fragments;

sending each of said plurality of data file fragments to be backed up to said subset of available target clients without tracking any location of where said plurality of data file fragments is sent; and

severing connections with said subset of said available target clients.

12. A method according to claim 10, wherein said trackless data backup-restore request is for restoring one or more data files, and said directing step further comprises the steps of:

selecting said one or more data files to be restored on said grid computing system;

establishing a predefined restoration level to restore said one or more data files selected;

transmitting said trackless data backup-restore request to said subset of said available target clients along with a respective trackless unique file identifier name corresponding to each of said one or more data files to be restored;

receiving communication from said subset of said available target clients that possess one or more data file fragments corresponding to said respective trackless unique file identifier name corresponding to each of said one or more data files;

retrieving said one or more data file fragments from said subset of said available target clients;

assembling said one or more data file fragments retrieved from said subset of said available clients to restore said one or more data files; and

severing connections with said subset of said available target clients.

13. A method according to claim 10, wherein said trackless data backup-restore request is for recreating said client node that has experienced said loss of data, and said directing step further comprises the steps of:

establishing a predefined restoration level to restore said client node;

recovering said client node to an initial base state;

restoring a unique identifier list file containing a list of all of a plurality of trackless unique file identifier names corresponding to all of a plurality of data files that said client node had backed up prior to said loss of data; and

employing said unique identifier list file to restore all of said plurality of data files that said client node had backed up prior to said loss of data.

14. A method according to claim 10, wherein said providing step further comprises the step of:

providing said cluster of client nodes that includes at least one super node configured to broker an initial communication regarding said trackless data backup-restore request between said source client and said plurality of target clients on said grid computing system, wherein said at least one super node communicates in said initial communication said respective trackless unique file identifier names corresponding to said one or more data files selected for said trackless backup-restore request, and wherein said at least one super node receives communication from said subset of said available target clients that possess said one or more data file fragments corresponding to said respective trackless unique file identifier names communicated.

15. A method according to claim 12, further comprising the steps of:

determining whether a sufficient number of said one or more data file fragments have been retrieved from said subset of said available target clients to restore each of said one or more data files selected to said predefined restoration level; and

if said sufficient number of said data file fragments have not been retrieved, initiating another data backup-restore request for acquiring additional data file fragments to restore each of said one or more data files selected to said predefined restoration level.

16. A method according to claim 13, wherein said employing step further comprises the steps of:

transmitting said trackless data backup-restore request to said subset of said available target clients along with said respective trackless unique file identifier names corresponding to said plurality of data files to be restored;

receiving communication from said subset of said available target clients that possess one or more data file fragments corresponding to said respective trackless unique file identifier names transmitted to restore said plurality of data files that said client had backed up prior to said loss of data;

retrieving said one or more data file fragments corresponding to said respective trackless unique file identifier names;

assembling each of said plurality of data files from said one or more data file fragments retrieved from said subset of said available target clients; and

severing connections with said subset of said available target clients.

17. A method according to claim 14, wherein said seeking step further comprises the steps of:

said at least one super node communicating said trackless data backup-restore request from said source client to said plurality of target clients and said at least one super node receiving communication from said available target clients regarding availability; and

said at least one super node supplying said source client with a list of said available target clients that are available for performing said trackless data backup-restore request.

18. A method of providing a trackless data backup-restoration service for backing up and restoring data on-demand across a plurality of peer nodes in a grid computing system, said method comprising the steps of:

providing said plurality of grid computing system peer nodes connected via a grid computing system network communications channel for said grid computing system, said plurality of grid computing system peer nodes having a grid computing software installed thereon for executing at least one grid application, said plurality of grid computing system peer nodes including a grid computing system source client configured to initiate a trackless data backup-restore request on-demand for one or more data files having one or more data file fragments, said one or more data files having respective trackless unique file identifier names, and a plurality of grid computing system target clients configured to perform said trackless data backup-restore request on-demand for said one or more data files having said respective trackless unique file identifier names;

supplying a trackless data backup-restore software to said plurality of grid computing system peer nodes for installation on said plurality of grid computing system peer nodes, said trackless data backup-restore software including a grid computing system super node list containing at least one grid computing system super node configured to broker said trackless data backup-restore request from said grid computing system source client to said plurality of grid computing system target clients;

communicating said trackless data backup-restore request, including said respective trackless unique file identifier names corresponding to said one or more data files, from said grid computing system source client to a respective grid computing system super node on said grid computing system super node list;

transmitting, by said respective grid computing system super node, said data backup-restore request from said source client to said plurality of target clients;

said respective grid computing system super node forwarding to said source client a grid computing system target list of available target clients that are available to perform said trackless data backup-restore request on-demand; and

establishing, by said source client, direct connections with said available target clients on said grid computing system target list for performing said trackless data backup-restore request on-demand, wherein said available target clients perform said trackless data backup-restore request for said one or more data file fragments on said grid computing system, without tracking a respective location of said one or more data file fragments backed up on or retrieved from said grid computing system.

19. A method according to claim 18, wherein said providing step further comprises the step of:

providing a subscription manager for registering said plurality of grid computing system peer nodes for participation in said trackless data backup-restoration service.

20. Programmable media storing programmable software to backup-restore data across a plurality of grid computing network peer nodes in a grid computing network, said plurality of grid computing network peer nodes having a grid computing software installed thereon for executing a grid application, said plurality of grid computing network peer nodes having installed thereon a trackless data backup-restore software for backing up and restoring data, said plurality of grid computing network peer nodes including a source client configured to initiate on-demand a trackless data backup-restore request for one or more data files fragmented into one or more data file fragments and a plurality of target clients configured to perform on-demand said trackless data backup-restore request for said one or more data file fragments, said programmable software comprising the steps of:

communicating said trackless data backup-restore request from said source client to said plurality of target clients, said trackless data backup-restore request including respective trackless unique file identifier names corresponding to said one or more data files;

establishing direct connections between said source client and a subset of available target clients that respond to said trackless data backup-restore request communicated; and

executing said trackless backup-restore request containing said respective trackless unique file identifier names corresponding to said one or more data files on said subset of available target clients without tracking respective locations of said data file fragments backed up on or retrieved from said subset of available target clients.

21. The programmable media according to claim 20, wherein said trackless data backup-restore request comprises one of:

backing up a data file, backing up one or more of a plurality of data file fragments, restoring a data file, retrieving a data file, retrieving one or more of a plurality of data file fragments, restoring one or more of a plurality of data file fragments and restoring a peer node of said plurality of peer nodes that has experienced a loss of data.

22. The programmable media according to claim 21, wherein said communicating step further comprises the step of:

providing a super node list that includes at least one super node configured to broker said trackless data backup-restore request on-demand from said source client to said plurality of target clients on said grid computing system, wherein said super node communicates said trackless data backup-restore request containing said respective trackless unique file identifier names corresponding to said one or more data files selected for said trackless data backup-restore request from said source client to said plurality of target clients and provides said source client with a list of said available target clients that responded to said trackless data backup-restore request communicated, and wherein said source client establishes direct connections with a subset of said available target clients for performing on-demand said trackless backup-restore request.