Data grid advisor

Patent metadata

Field	Value
Publication number	US-9529846-B2
Application number	US-97003910-A
Country	US
Kind code	B2
Filing date	Dec 16, 2010
Priority date	Dec 16, 2010
Publication date	Dec 27, 2016
Grant date	Dec 27, 2016

Abstract

A system and method to generate an improved layout of a data grid in a database environment is provided. The data grid is a clustered in-memory database cache comprising one or more data fabrics, where each data fabric includes multiple in-memory database cache nodes. A data grid advisor capability can be used by application developers and database administrators to evaluate and design the data grid layout so as to optimize performance based on resource constraints and the needs of particular database applications.

First claim

What is claimed is:

1. A computer-implemented method to produce an improved layout of a data grid in a database environment comprising:
capturing, by a computing device, a workload from a first row-oriented database server and at least one client of the first database server, the workload comprising a set of queries and responses between the first database server and the at least one client;
producing dependency and volume information for the captured workload based on the set of queries and responses in the captured workload;
generating a layout of one or more data fabrics within the data grid in the database environment based on resource constraints associated with the data grid and the produced dependency and volume information associated with the captured workload, wherein each fabric in the one or more data fabrics comprises a plurality of cache nodes, and wherein each cache node in the plurality of cache nodes is an in-memory database server; and
storing the generated layout at the computing device.

2. The method of claim 1 , further comprising:
identifying a subset of queries from the set of queries in the captured workload based on the resource constraints associated with the data grid and the produced dependency and volume information associated with the captured workload, wherein one or more query-processing latency reduction techniques may be used for the subset of queries at runtime; and
generating a workset for the data grid in the database environment, the workset comprising the identified subset of queries.

3. The method of claim 1 , wherein the producing comprises:
replaying the set of queries and responses in the captured workload using a second database server to produce the dependency and volume information corresponding to the set of queries and responses in the captured workload, wherein the second database server is a replica of the first database server.

4. The method of claim 1 ,
wherein the capturing comprises:
simulating, by the computing device, the queries and responses from the first database server to the at least one client based on user input and at least one database schema,

wherein the producing comprises:
producing synthetic dependency and volume information corresponding to the simulated queries and responses based on the simulating, and

wherein the generating comprises:
generating the layout of the one or more data fabrics within the data grid in the database environment based on the resource constraints associated with the data grid and the synthetic dependency and volume information.

5. The method of claim 1 , wherein the resource constraints associated with the data grid comprise hardware resource limits, at least one database schema, and a type of data granularity.

6. The method of claim 5 , wherein the type of data granularity is database granularity, and wherein an entire backend database in the database environment is copied to the one or more data fabrics in the data grid.

7. The method of claim 5 , wherein the type of data granularity is partition granularity, and wherein the generating further comprises:
partitioning data from one or more tables in a backend database into data slices;
distributing the data slices within the one or more data fabrics of the data grid based on the hardware resource limits and the at least one database schema, wherein a single cache node from the one or more cache nodes is selected to have exclusive read-write ownership of the partitioned data.

8. A non-transitory computer readable storage medium having instructions stored thereon that, when executed by a computing device, cause the computing device to perform operations comprising:
capturing a workload from a first row-oriented database server and at least one client of the first database server, the workload comprising a set of queries and responses between the first database server and the at least one client;
producing dependency and volume information for the captured workload based on the set of queries and responses in the captured workload;
generating a layout of one or more data fabrics within a data grid in the database environment based on resource constraints associated with the data grid and the produced dependency and volume information associated with the captured workload, wherein each fabric in the one or more data fabrics comprises a plurality of cache nodes, and wherein each cache node in the plurality of cache nodes is an in-memory database server; and
storing the generated layout at the computing device.

9. The computer readable storage medium of claim 8 , the operations further comprising:
identifying a subset of queries from the set of queries in the captured workload based on the resource constraints associated with the data grid and the produced dependency and volume information associated with the captured workload, wherein one or more query-processing latency reduction techniques may be used for the subset of queries at runtime; and
generating a workset for the data grid in the database environment, the workset comprising the identified subset of queries.

10. The computer readable storage medium of claim 8 , wherein the producing comprises:
replaying the set of queries and responses in the captured workload using a second database server to produce the dependency and volume information corresponding to the set of queries and responses in the captured workload, wherein the second database server is a replica of the first database server.

11. The computer readable storage medium of claim 8 ,
wherein the capturing comprises:
simulating, by the computing device, the queries and responses from the first database server to the at least one client based on user input and at least one database schema,

wherein the producing comprises:
producing synthetic dependency and volume information corresponding to the simulated queries and responses based on the simulating, and

wherein the generating comprises:
generating the layout of data and applications for the data grid in the database environment based on the resource constraints associated with the data grid and the synthetic dependency and volume information.

12. The computer readable storage medium of claim 8 , wherein the resource constraints associated with the data grid comprise hardware resource limits, at least one database schema, and a type of data granularity.

13. The computer readable storage medium of claim 12 , wherein the type of data granularity is database granularity, and wherein an entire backend database in the database environment is copied to the one or more data fabrics in the data grid.

14. The computer readable storage medium of claim 12 , wherein the type of data granularity is partition granularity, and wherein the generating further comprises:
partitioning data from one or more tables in a backend database into data slices;
distributing the data slices within the one or more data fabrics of the data grid based on the hardware resource limits and the at least one database schema, wherein a single cache node from the one or more cache nodes is selected to have exclusive read-write ownership of the partitioned data.

15. A system to produce an improved layout of a data grid in a database environment comprising:
a workload preparer to capture a workload from a first row-oriented database server and at least one client of the first database server, the workload comprising a set of queries and responses between the first database server and the at least one client, and to produce dependency and volume information for the captured workload based on the set of quer

Assignees

• Wu Juan
• Andre Mihnea
• Du Haiyan
• Macleod Ian Scott
• Sybase Inc

Inventors

• Wu Juan
• Andre Mihnea
• Du Haiyan
• Macleod Ian Scott

Classifications

• G06F16/24532 Primary

  of parallel queries · CPC title

• G06F16/2471

  Distributed queries · CPC title

• G06F17/30545

  Physics · mapped topic

• G06F17/30445 Primary

  Physics · mapped topic