- 专利标题: Intelligent memory management through peer learning
-
申请号: US15973885申请日: 2018-05-08
-
公开(公告)号: US10671432B2公开(公告)日: 2020-06-02
- 发明人: Daniel M. Hursh , Jonathan L. Kaus , David M. Koster , Jason A. Nikolai
- 申请人: International Business Machines Corporation
- 申请人地址: US NY Armonk
- 专利权人: International Business Machines Corporation
- 当前专利权人: International Business Machines Corporation
- 当前专利权人地址: US NY Armonk
- 代理机构: Winstead PC
- 代理商 Robert A. Voigt, Jr.
- 主分类号: G06F9/50
- IPC分类号: G06F9/50 ; G06F3/06 ; G06F9/48 ; G06N20/00
摘要:
A method, system and computer program product for optimizing memory management. An analytics engine receives from a computing device an execution sequence that led to the detected execution of an operation to create a new process, which involves copying of the parent process. The analytics engine searches profiles stored in a repository for patterns of execution sequences that match the received execution sequence. Upon identifying a stored pattern with an execution sequence within a threshold degree of similarity as the received execution sequence, the analytics engine instructs the computing device to handle the operation to create the new process in a certain manner based on the contents of the profile containing the matching stored pattern. For example, the computing device may be instructed to preemptively fail the operation in response to the profile indicating that the execution of the operation resulted in an out of memory error.
信息查询