Data access management in a hybrid memory server
AssigneeInternational Business Machines Corporation
InventorsYuk Lung Chan, Rajaram B. Krishnamurthy, Carl Joseph Parris
Life of this patent
Estimating the term…
Abstract
A method, accelerator system, and computer program access data in an out-of-core processing environment. A data access configuration is received from a server system managing a plurality of data sets. A determination is made that data sets retrieved from the server system are to be stored locally based on the data access configuration. A request to interact with a given data set is received from a user client. At least a portion of the given data set is retrieved from the server system. The at least a portion of the given data set is stored locally a memory based on the data access configuration that has been received.
Claims15
A method, on an accelerator processing system comprising one or more processing cores, for accessing data in an out-of-core processing environment, the accelerator processing system being separate and distinct from at least one user client system and at least one server system managing a plurality of data sets stored remotely from the accelerator processing system, the method comprising:
- receiving a data access configuration from the server system managing the plurality of data sets, the data access configuration dynamically configuring how the accelerator processing system accesses and caches one or more of the plurality of data sets based on an access context associated with one or more of the plurality of data sets, the access context being based on at least one of a set of attributes associated with a user requesting access to the one or more of the plurality of datasets, a set of security attributes associated with the one or more of the plurality of datasets;
- a first set of data ports being used between the server system and the accelerator system, and a second set of data ports being used between the accelerator system and the user client system;
- determining, based on the data access configuration, that data sets retrieved from the server system are to be stored locally;
- receiving, from the user client system, a request to interact with a given data set in the plurality of data sets;
- retrieving at least a portion of the given data set from the server system;
- and
- storing the at least a portion of the given data set locally in a memory based on the data access configuration that has been received.
An accelerator processing system in an out-of-core processing environment comprising the accelerator processing system, at least one server system managing a plurality of data sets stored remotely from the accelerator processing system, and at least one user client system, the accelerator processing system being separate and distinct from the server system and the user client system, the accelerator system comprising:
- a memory;
- a processing core communicatively coupled to the memory;
- and
- a data manager communicatively coupled to the memory and the processing core, wherein the data manager is configured to perform a method comprising:
- receiving a data access configuration from the server system managing the plurality of data sets, the data access configuration dynamically configuring how the accelerator processing system accesses and caches one or more of the plurality of data sets based on an access context associated with one or more of the plurality of data sets, the access context being based on at least one of a set of attributes associated with a user requesting access to the one or more of the plurality of datasets, a set of security attributes associated with the one or more of the plurality of datasets;
- a first set of data ports being used between the server system and the accelerator system, and a second set of data ports being used between the accelerator system and the user client system;
- determining, based on the data access configuration, that data sets retrieved from the server system are to be stored locally;
- receiving, from the user client system, a request to interact with a given data set in the plurality of data sets;
- retrieving at least a portion of the given data set from the server system;
- and
- storing the at least a portion of the given data set locally in a memory based on the data access configuration that has been received.
A computer program product for accessing data in an out-of-core processing environment, the out-of-core processing environment comprising at least one accelerator processing system, at least one server system managing a plurality of data sets stored remotely from the at least one accelerator processing system, and at least one user client system, the at least one accelerator processing system being separate and distinct from the at least one server system, the computer program product comprising:
- a non-transitory storage medium readable by a processing circuit and storing instructions for execution by the processing circuit for performing a method, by the at least one accelerator processing system, comprising:
- receiving a data access configuration from the server system managing a plurality of data sets, the data access configuration dynamically configuring how the accelerator processing system accesses and caches one or more of the plurality of data sets based on an access context associated with one or more of the plurality of data sets, the access context being based on at least one of a set of attributes associated with a user requesting access to the one or more of the plurality of datasets, a set of security attributes associated with the one or more of the plurality of datasets;
- a first set of data ports being used between the server system and the accelerator system, and a second set of data ports being used between the accelerator system and the user client system;
- determining, based on the data access configuration, that data sets retrieved from the server system are to be stored locally;
- receiving, from the user client system, a request to interact with a given data set in the plurality of data sets;
- retrieving at least a portion of the given data set from the server system;
- and
- storing the at least a portion of the given data set locally in a memory based on the data access configuration that has been received.
Dependencies are inferred from the claim text.