System for distributed data processing with auto-recovery

• US 9,952,942 B2
• Filed: 02/12/2016
• Issued: 04/24/2018
• Est. Priority Date: 02/12/2016
• Status: Active Grant

• Current Assignee: Bank of America Corp.
• Original Assignee: Bank of America Corp.
• Inventors: Gunsolley, Shawn Cart, Cassell, Erin, Potla, Siva Shankar, Desautels, Adam Nathaniel, Poore, Jeffrey Scott, Thompson, Marshall Bright
• Primary Examiner(s): Manoskey, Joseph D
• Application Number:
  US15/042,290
• Publication Number:
  US 20170235610A1
• Time in Patent Office: 802 Days
• Field of Search: 714 15, 714 20, 714 16
• CPC Class Codes: G06F 11/0709 in a distributed system consisting of a plurality of standalone computer nodes, e.g. clusters, client-server systems
• CPC Class Codes: G06F 11/1438 Restarting or rejuvenating
• CPC Class Codes: G06F 11/1458 Management of the backup or restore process
• CPC Class Codes: G06F 11/1471 involving logging of persistent data for recovery
• CPC Class Codes: G06F 11/2035 without idle spare hardware
• CPC Class Codes: G06F 9/461 Saving or restoring of program or task context
• CPC Class Codes: G06F 9/5027 the resource being a machine, e.g. CPUs, Servers, Terminals
• CPC Class Codes: G06F 9/5044 considering hardware capabilities
• CPC Class Codes: G06F 9/52 Program synchronisation; Mutual exclusion, e.g. by means of semaphores
• CPC Class Codes: G06Q 30/04 Billing or invoicing

Abstract

Embodiments enable distributed data processing with automatic caching at multiple system levels by accessing a master queue of data processing work comprising a plurality of data processing jobs stored in a long term memory cache; selecting at least one of the plurality of data processing jobs from the master queue of data processing work; pushing the selected data processing jobs to an interface layer including (i) accessing the selected data processing jobs from the long term memory cache; and (ii) saving the selected data processing jobs in an interface layer cache of data processing work; and pushing at least a portion of the selected data processing jobs to a memory cache of a first user system for minimizing latency in user data processing of the pushed data processing jobs.

Claims

1. A data processing system for distributed data processing, the data processing system comprising:

a memory device with computer-readable program code stored thereon;

a communication device;

a processing device operatively coupled to the memory device and the communication device, wherein the processing device is configured to execute the computer-readable program code to;

access a master queue of data processing work comprising a plurality of data processing jobs stored in a long term memory cache;

select at least one of the plurality of data processing jobs from the master queue of data processing work;

divide the at least one data processing job into a plurality of data processing items;

allocate each of the plurality of data processing items to a different one of a distributed network comprising a plurality of distributed user systems to ensure maximum efficiency in processing the at least one data processing job;

actively synchronize some or all the plurality of data processing items among the plurality of distributed user systems, the actively synchronizing comprising;

repeatedly or periodically saving results of the data processing; and

processing the data processing items at a smallest block level allowed by each of the distributed user systems, thereby maximizing efficiency of automatic recovery of completed data processing work.

2. The data processing system of claim 1, wherein the processing device is further configured to execute computer-readable code to:

store the data processing jobs in an in-flight data table; and

monitor the in-flight data table to ensure efficient data processing of the data processing job.

3. The data processing system of claim 1, wherein actively synchronizing the data processing items comprises repeatedly or periodically saving a status of progress of data processing.

4. The data processing system of claim 1, wherein the processing device is further configured to execute computer-readable code to:

automatically recover, in response to a processing fault by one of the user systems, completed data processing work using the saved results of the data processing.

5. The data processing system of claim 1, wherein the repeated or periodic saving of results of the data processing comprises saving some or all the results on the distributed network of user systems.

6. The data processing system of claim 1, wherein the processing device is further configured to execute computer-readable code to:

determine a work capacity for each of the user systems on the distributed network; and

wherein allocating is based on the determined work capacity for each of the user systems.

7. A computer program product for distributed data processing, the computer program product comprising at least one non-transitory computer-readable medium having computer-readable program code portions embodied therein, the computer-readable program code portions comprising:

an executable portion configured for accessing a master queue of data processing work comprising a plurality of data processing jobs stored in a long term memory cache;

an executable portion configured for selecting at least one of the plurality of data processing jobs from the master queue of data processing work;

an executable portion configured for dividing the at least one data processing job into a plurality of data processing items;

an executable portion configured for allocating each of the plurality of data processing items to a different one of a distributed network comprising a plurality of distributed user systems to ensure maximum efficiency in processing the at least one data processing job;

an executable portion configured for actively synchronizing some or all the plurality of data processing items among the plurality of distributed user systems, the actively synchronizing comprising;

repeatedly or periodically saving results of the data processing; and

processing the data processing items at a smallest block level allowed by each of the distributed user systems, thereby maximizing efficiency of automatic recovery of completed data processing work.

8. The computer program product of claim 7, wherein the computer-readable program code portions further comprise:

an executable portion configured for storing the data processing jobs in an in-flight data table; and

an executable portion configured for monitoring the in-flight data table to ensure efficient data processing of the data processing job.

9. The computer program product of claim 7, wherein actively synchronizing the data processing items comprises repeatedly or periodically saving a status of progress of data processing.

10. The computer program product of claim 7, wherein the computer-readable program code portions further comprise:

an executable portion configured for automatically recovering, in response to a processing fault by one of the user systems, completed data processing work using the saved results of the data processing.

11. The computer program product of claim 7, wherein the repeated or periodic saving of results of the data processing comprises saving some or all the results on the distributed network of user systems.

12. The computer program product of claim 7, wherein the computer-readable program code portions further comprise:

an executable portion configured for determining a work capacity for each of the user systems on the distributed network; and

wherein allocating is based on the determined work capacity for each of the user systems.

13. A computer-implemented method for distributed data processing, the method comprising:

accessing a master queue of data processing work comprising a plurality of data processing jobs stored in a long term memory cache;

selecting at least one of the plurality of data processing jobs from the master queue of data processing work;

dividing the at least one data processing job into a plurality of data processing items;

allocating each of the plurality of data processing items to a different one of a distributed network comprising a plurality of distributed user systems to ensure maximum efficiency in processing the at least one data processing job; and

actively synchronizing some or all the plurality of data processing items among the plurality of distributed user systems, the actively synchronizing comprising;

repeatedly or periodically saving results of the data processing; and

processing the data processing items at a smallest block level allowed by each of the distributed user systems, thereby maximizing efficiency of automatic recovery of completed data processing work.

14. The method of claim 13, further comprising:

storing the data processing jobs in an in-flight data table; and

monitoring the in-flight data table to ensure efficient data processing of the data processing job.