发明名称 Optimized transfer and storage of highly denormalized data in an in-memory data grid
摘要 A method for transferring and storing a data set associated with an in-memory data grid (IMDG) is provided. The method may include identifying at least two data object instances associated with the data set based on a determined relationship between the at least two data object instances. The method may also include performing a field relation analysis on the identified at least two data object instances and recording the field relation analysis. The method may include receiving a request for a data object associated with the data set. The method may include determining if an optimization can be performed on the data object. The method may include omitting a field value associated with the data object based on the determined optimization. The method may include inflating the omitted field value into the data set associated with the data object request and returning the data set with the inflated field value.
申请公布号 US9329786(B2) 申请公布日期 2016.05.03
申请号 US201414526809 申请日期 2014.10.29
申请人 International Business Machines Corporation 发明人 Amrhein Dustin K.;Berg Douglas C.;Bhogal Kulvir S.;Gaur Nitin;Johnson Christopher D.
分类号 G06F17/30;G06F3/06 主分类号 G06F17/30
代理机构 代理人 Feighan Patricia B.
主权项 1. A method for transferring and storing a data set associated with an in-memory data grid (IMDG), the method comprising: in response to the data set being loaded into the IMDG associated with a first computer, identifying at least two data object instances associated with the loaded data set based on a determined relationship between the at least two data object instances, wherein the identifying occurs periodically; performing a field relation analysis on the identified at least two data object instances, wherein performing a field relation analysis comprises determining whether the identified at least two data object instances share a reference field value; prompting a user using a second computer to determine whether to optimize the identified at least two data instances sharing a reference field value; in response to the user determining to optimize the identified at least two data instances sharing a reference field value, recording the field relation analysis; receiving a request for a data object associated with the data set; determining if an optimization can be performed on the data object based on the recorded field relation analysis; omitting a field value associated with the data object based on the determined optimization; inflating the omitted field value into the data set associated with the data object request; returning the data set with the inflated field value; and recording a plurality of metadata associated with the received request, wherein the recorded metadata comprises a time the request was received, a time the data set was returned, an object type associated with the requested data object, and an ID associated with the requested data object.
地址 Armonk NY US