Autonomic data partition placement in an in-memory data grid
AssigneeInternational Business Machines Corporation
InventorsDustin K. Amrhein, Douglas C. Berg, Nitin Gaur, Christopher D. Johnson
Life of this patent
Estimating the term…
Abstract
A method for autonomic data partition placement in an in-memory data grid (IMDG) based on usage sampling is provided. The method may include sampling access pattern data, during a configurable window, of a client accessing one or more data objects in the IMDG. Analytics are applied to the sampled access pattern data to determine a replacement partition schema, based on the analyzed sampled access pattern data exceeding at least one performance threshold. The deployed partition schema is altered, based on the replacement partition schema.
Claims12
A method for autonomic data partition placement in an in-memory data grid (IMDG) based on usage sampling comprising:
- sampling access pattern data, during a configurable window, of a client accessing one or more data objects in the IMDG, wherein access pattern data comprises:
- 1) a number of read operations performed, 2) a number of write operations performed, 3) a frequency by which the one or more data objects in the IMDG is accessed, 4) an amount of data transferred during a read operation, and 5) an amount of data transferred during a write operation;
- applying analytics to the sampled access pattern data to determine a replacement partition schema, based on the analyzed sampled access pattern data exceeding at least one performance threshold;
- and
- altering a deployed partition schema, based on the replacement partition schema, wherein the altering is transparent to the client accessing the one or more data objects in the IMDG.
A computer program product for autonomic data partition placement in an in-memory data grid (IMDG) based on usage sampling comprising:
- a computer readable storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising:
- sampling access pattern data, during a configurable window, of a client accessing one or more data objects in the IMDG, wherein access pattern data comprises:
- 1) a number of read operations performed, 2) a number of write operations performed, 3) a frequency by which the one or more data objects in the IMDG is accessed, 4) an amount of data transferred during a read operation, and 5) an amount of data transferred during a write operation;
- applying analytics to the sampled access pattern data to determine a replacement partition schema, based on the analyzed sampled access pattern data exceeding at least one performance threshold, wherein the at least one performance threshold comprises:
- 1) a maximum percentage of CPU usage, 2) a maximum percentage of memory usage, 3) exceeding a limit of allocated storage by at least one of the data objects in the IMDG, 4) a response time, and 5) exceeding a number of accesses to at least one of the data objects in the IMDG, based on the data object in the IMDG being accessed at a statistically higher rate than other data objects in the IMDG;
- and
- altering a deployed partition schema, based on the replacement partition schema, wherein the altering is transparent to the client accessing the one or more data objects in the IMDG.
A computer system for autonomic data partition placement in an in-memory data grid (IMDG) based on usage sampling, comprising one or more processors, one or more computer-readable storage devices, and a plurality of program instructions stored on at least one of the one or more storage devices for execution by at least one of the one or more processors, the plurality of program instructions comprising:
- a computer readable storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method comprising:
- sampling access pattern data, during a configurable window, of a client accessing one or more data objects in the IMDG, wherein access pattern data comprises:
- 1) a number of read operations performed, 2) a number of write operations performed, 3) a frequency by which the one or more data objects in the IMDG is accessed, 4) an amount of data transferred during a read operation, and 5) an amount of data transferred during a write operation;
- applying analytics to the sampled access pattern data to determine a replacement partition schema, based on the analyzed sampled access pattern data exceeding at least one performance threshold, wherein the at least one performance threshold comprises:
- 1) a maximum percentage of CPU usage, 2) a maximum percentage of memory usage, 3) exceeding a limit of allocated storage by at least one of the data objects in the IMDG, 4) a response time, and 5) exceeding a number of accesses to at least one of the data objects in the IMDG, based on the data object in the IMDG being accessed at a statistically higher rate than other data objects in the IMDG;
- and
- altering a deployed partition schema, based on the replacement partition schema, wherein the altering is transparent to the client accessing the one or more data objects in the IMDG.
Dependencies are inferred from the claim text.