公开(公告)号：US09652291B2

公开(公告)日：2017-05-16

申请号：US15173526

申请日：2016-06-03

Applicant: Palantir Technologies Inc.

Inventor： Punya Biswal ,  Beyang Liu ,  Eugene Marinelli ,  Nima Ghamsari

IPC: G06F12/08 ,  G06F9/50 ,  G06F17/30 ,  G06F12/084 ,  G06F12/0813 ,  G06F12/1009

CPC classification number: G06F9/5016 ,  G06F12/0813 ,  G06F12/084 ,  G06F12/1009 ,  G06F17/3048 ,  G06F2212/1016 ,  G06F2212/60 ,  G06F2212/65

Abstract: Methods and systems for providing a plurality of applications with concurrent access to data are disclosed. One such method includes identifying attributes of an expected data set to be accessed concurrently by the applications, initializing a shared cache with a column data store configured to store the expected data set in columns, and creating a memory map for accessing a physical memory location in the shared cache. The method may also include mapping the applications' data access requests to the shared cache with the memory map. Only one instance of the expected data set is stored in memory, so each application is not required to create additional instances of the expected data set in the application's memory address space. Therefore, larger expected data sets may be entirely stored in memory without limiting the number of applications running concurrently.

公开(公告)号：US09367463B2

公开(公告)日：2016-06-14

申请号：US13827627

申请日：2013-03-14

Applicant: Palantir Technologies, Inc.

Inventor： Punya Biswal ,  Beyang Liu ,  Eugene Marinelli ,  Nima Ghamsari

IPC: G06F12/08 ,  G06F17/30 ,  G06F9/50

CPC classification number: G06F9/5016 ,  G06F12/0813 ,  G06F12/084 ,  G06F12/1009 ,  G06F17/3048 ,  G06F2212/1016 ,  G06F2212/60 ,  G06F2212/65

Abstract: Methods and systems for providing a plurality of applications with concurrent access to data are disclosed. One such method includes identifying attributes of an expected data set to be accessed concurrently by the applications, initializing a shared cache with a column data store configured to store the expected data set in columns and creating a memory map for accessing a physical memory location in the shared cache. The method may also include mapping the applications' data access requests to the shared cache with the memory map. Only one instance of the expected data set is stored in memory, so each application is not required to create additional instances of the expected data set in the application's memory address space. Therefore, larger expected data sets may be entirely stored in memory without limiting the number of applications running concurrently.

Abstract translation: 公开了用于提供对数据的并发访问的多个应用的​​方法和系统。 一种这样的方法包括识别由应用程序同时访问的期望数据集的属性，使用被配置为在列中存储期望数据集的列数据存储器初始化共享高速缓存，并且创建存储器映射以访问存储器中的物理存储器位置 共享缓存。 该方法还可以包括使用存储器映射将应用的数据访问请求映射到共享高速缓存。 只有预期数据集的一个实例存储在内存中，因此每个应用程序不需要在应用程序的存储器地址空间中创建预期数据集的附加实例。 因此，较大的预期数据集可以完全存储在存储器中，而不会限制同时运行的应用程序的数量。