Dedicated fallback processing for a distributed data warehouse

Patent metadata

Field	Value
Publication number	US-10776229-B2
Application number	US-201715852654-A
Country	US
Kind code	B2
Filing date	Dec 22, 2017
Priority date	Dec 22, 2017
Publication date	Sep 15, 2020
Grant date	Sep 15, 2020

Abstract

Database processing engines of a single cluster are configured such that each engine is a primary engine and a dedicated fallback engine to one other engine of the cluster. In an embodiment, the cluster includes more than two processing engines.

First claim

The invention claimed is:

1. A method for dedicated fallback processing within a distributed data warehouse, said distributed data warehouse including a plurality of processing units arranged in a plurality of clusters, said method comprising:
configuring first, second, and third processing units within a single cluster;
configuring the first processing unit within the single cluster to process as a first dedicated fallback processing unit for the second processing unit of the single cluster;
configuring the second processing unit as a second dedicated fallback processing unit for the third processing unit of the single cluster; and
configuring the third processing unit as a third dedicated fallback processing unit for the first processing unit.

2. The method of claim 1 further comprising, processing, by the first processing unit, first operations against a first set of tables assigned to the first processing unit.

3. The method of claim 2 , wherein processing further includes providing, by the first processing unit, data for the first operations to the third processing unit acting as the third dedicated fallback processing unit.

4. The method of claim 3 , wherein providing further includes taking over processing of the first operations, by the third processing unit using the data, when the first processing unit becomes unresponsive.

5. The method of claim 2 , wherein processing further includes receiving data, by the first processing unit, from the second processing unit when the second processing unit processes second operations against a second set of tables assigned to the second processing unit with the first processing unit actions as the first dedicated fallback processing unit.

6. The method of claim 5 , wherein receiving further includes taking over processing of the second operations, by the first processing unit using the data, when the second processing unit becomes unresponsive.

7. The method of claim 1 , further comprising, processing, by the second processing unit, second operations against a second set of tables assigned to the second processing unit.

8. The method of claim 7 , wherein processing further includes receiving, by the second processing unit, data from the third processing unit when the third processing unit processes third operations against a third set of tables assigned to the third processing unit.

9. The method of claim 8 , wherein receiving further includes taking over processing of the third operations, by the second processing unit using the data when the third processing unit becomes unresponsive.

10. The method of claim 1 further comprising, processing by the third processing unit, third operations against a third set of tables assigned to the third processing unit.

11. The method of claim 10 , wherein processing further includes receiving data from the first processing unit when the first processing unit processes first operations against a first set of tables assigned to the first processing unit.

12. A method for dedicated fallback processing within a distributed data warehouse, said distributed data warehouse including a plurality of processing units arranged in a plurality of clusters, said method comprising:
configuring a sub-plurality of processing units within a single cluster;
processing each one of the processing units within the sub-plurality of processing units in the single cluster of the distributed data warehouse as a primary processing unit and as a fallback processing unit to a select one of remaining processing units within the sub-plurality of processing units within the single cluster.

13. The method of claim 12 further comprising, processing each of the fallback processing units when a corresponding one of the primary processing units becomes unresponsive.

14. The method of claim 12 further comprising, providing from each of the primary processing units data associated with processing operations against tables assigned to that primary processing unit to a corresponding one of the fallback processing units.

15. The method of claim 12 , wherein the sub-plurality of processing units comprises more than two processing units.

16. The method of claim 12 , wherein processing further includes processing the plurality of processing units as an even number of processing units.

17. The method of claim 12 , wherein processing further includes processing the plurality of processing units as an odd number of processing units.

18. The method of claim 12 , wherein processing further includes assigning a unique set of tables for processing by each unique processing unit.

19. A system, comprising:
a distributed data warehouse including:
a plurality of processing units arranged in a plurality of clusters;
a sub-plurality of processing units within a single cluster;
a dedicated fallback configuration manager;
wherein the dedicated fallback configuration manager is configured to: i) execute on at least one network node of the data warehouse ii) configure and process each one of the processing units within the sub-plurality of processing units within the single cluster of the data warehouse as a primary processing unit and as a dedicated fallback processing unit for a unique one of remaining ones of the processing units within the sub-plurality of processing units within the single cluster.

20. The system of claim 19 . wherein a total number of processing units within the sub-plurality of processing units within the single cluster is greater than two and the total number is one of: an even number and an odd number.

Assignees

• Teradata Us Inc

Inventors

• Krishna Tirunagari Venkata Rama
• Boppudi Srinath
• Kaufmann Frederick Stuart
• Pederson Donald Raymond

Classifications

• G06F11/2048

  where the redundant components share neither address space nor persistent storage · CPC title

• G06F9/4881 Primary

  Scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues · CPC title

• G06F11/2025

  using centralised failover control functionality · CPC title

• G06F11/2097 Primary

  maintaining the standby controller/processing unit updated (initialisation or re-synchronisation thereof G06F11/1658 and subgroups) · CPC title

• G06F2201/80

  Database-specific techniques · CPC title