Hybrid OLTP and OLAP High Performance Database System

• US 20130073513A1
• Filed: 04/04/2011
• Published: 03/21/2013
• Est. Priority Date: 05/17/2010
• Status: Active Grant

• Current Assignee: Technische Universitat Munchen
• Original Assignee: Technische Universitat Munchen
• Inventors: Kemper, Alfons, Thomas, Neumann
• Granted Patent: US 10,002,175 B2
• Time in Patent Office: Days
• US Class Current: 707/600
• CPC Class Codes: G06F 16/28 Databases characterised by their database models, e.g. relational or object models

Abstract

There is provided a method of maintaining a hybrid OLTP and OLAP database, the method comprising: executing one or more OLTP transactions; creating a virtual memory snapshot; and executing one or more OLAP queries using the virtual memory snapshot. Preferably, the method further comprises replicating a virtual memory page on which a data object is stored in response to an update to the data object, whereby the updated data object is accessible for OLTP transactions, while the non-updated data object remains accessible for OLAP queries. Accordingly, the present invention provides a hybrid system that can handle both OLTP and OLAP simultaneously by using hardware-assisted replication mechanisms to maintain consistent snapshots of the transactional data.

Claims

1. A method of maintaining a hybrid OLTP and OLAP database system comprising a memory, the method comprising:

executing one or more OLTP transactions;

creating one or more virtual memory snapshots; and

executing one or more OLAP queries using one or more of the virtual memory snapshots.

2. The method of claim 1, further comprising replicating a virtual memory page on which a data object is stored in response to an update to the data object, whereafter the updated data object is accessible for OLTP transactions, while the non-updated data object remains accessible for OLAP queries.

3. The method of claim 1, wherein the OLTP transactions are executed in a first address space, wherein creating the virtual memory snapshots comprises providing a second address space that represents a copy of the first address space, wherein the OLAP queries are executed in the second address space, and wherein creating the virtual memory snapshots comprises duplicating the OLTP process to create a child process for execution of OLAP queries.

4. The method of claim 1, wherein the database stores private data in partitions and shared data, wherein the partitions may reside on different data processing systems, the method comprising:

executing, in parallel, OLTP transactions that comprise read and/or update accesses to the private data or read accesses to the shared data;

executing, in sequence, OLTP transactions that comprise read accesses across the partitions or update accesses to the shared data; and

/or

executing one or more OLAP queries across one ore more of the partitions and the shared data.

5. The method of claim 1, wherein the virtual memory snapshots are transaction-consistent.

6. The method of claim 5, further comprising creating the virtual memory snapshots when no OLTP transaction is active.

7. The method of claim 5, further comprising creating the virtual memory snapshots when one or more OLTP transactions are active, and using an undo-log mechanism to adapt the virtual memory snapshots to represent the state of the database before the one or more active OLTP transactions were initiated.

8. The method of claim 1, further comprising executing, in parallel, multiple OLAP queries using one of the virtual memory snapshots.

9. The method of claim 1, further comprising creating multiple virtual memory snapshots for respective parallel OLAP queries.

10. The method of claim 1, further comprising creating new virtual memory snapshots periodically or on demand.

11. The method of claim 1, further comprising using an existing hardware-supported memory consistency control mechanism to identify when a data object in the database is updated and to trigger the creation of a new physical copy of the corresponding page.

12. The method of claim 1, further comprising deleting the virtual memory snapshot after completion of a corresponding OLAP query.

13. The method of claim 1, further comprising using the virtual memory snapshots to provide a transaction-consistent backup of the database.

14. The method of claim 13, further comprising using the transaction-consistent backup and a redo log mechanism to restore the database.

15. The method of claim 1, further comprising

maintaining a primary server to execute said OLTP transactions, and

maintaining a secondary server to also execute at least a portion of the OLTP transactions executed by the primary server, in particular all OLTP transactions comprising update accesses to the data, and to execute at least some of said OLAP queries.

16. A hybrid OLTP and OLAP database system, comprising:

a database stored in a memory;

a processor; and

a storage, coupled to the processor, storing instructions that, when executed, cause the processor to perform the method of claim 1.

17. The hybrid OLTP and OLAP database system of claim 16, wherein the database is a main memory database.

18. An article comprising a medium storing instructions that, when executed, cause a processor-based system to perform the method of claim 1.

19. (canceled)