公开(公告)号：US20150113091A1

公开(公告)日：2015-04-23

申请号：US14061515

申请日：2013-10-23

Applicant: Yahoo! Inc.

Inventor： Amarjit Luwang Thiyam ,  Saurabh Singla

IPC: H04L29/08

CPC classification number: H04L67/2842 ,  G06F12/0815 ,  G06F2212/264 ,  G06F2212/284

Abstract: In an example of masterless cache replication, a processor of a server of a plurality of servers hosting a distributed application can receive a local cache event for a local data item stored in an application cache of the server. The processor can determine whether the local cache event is from another server. The processor can also determine whether a remote cache event of the other server is different from the local cache event and whether the local cache event is in conflict with at least one other cache event for the local data item. The processor can also determine whether the local cache event has a higher priority over the at least one other cache event and direct performance of the local cache event amongst the plurality of servers.

Abstract translation: 在无主缓存复制的示例中，托管分布式应用的多个服务器的服务器的处理器可以接收存储在服务器的应用高速缓存中的本地数据项的本地高速缓存事件。 处理器可以确定本地缓存事件是否来自另一个服务器。 处理器还可以确定其他服务器的远程高速缓存事件是否不同于本地高速缓存事件以及本地高速缓存事件是否与本地数据项的至少一个其他高速缓存事件冲突。 处理器还可以确定本地高速缓存事件是否比该至少一个其他高速缓存事件具有更高的优先级，并且在多个服务器中直接执行本地高速缓存事件。

公开(公告)号：US09602615B2

公开(公告)日：2017-03-21

申请号：US14061515

申请日：2013-10-23

Applicant: Yahoo! Inc.

Inventor： Amarjit Luwang Thiyam ,  Saurabh Singla

IPC: H04L29/08 ,  G06F12/08

CPC classification number: H04L67/2842 ,  G06F12/0815 ,  G06F2212/264 ,  G06F2212/284

Abstract: In an example of masterless cache replication, a processor of a server of a plurality of servers hosting a distributed application can receive a local cache event for a local data item stored in an application cache of the server. The processor can determine whether the local cache event is from another server. The processor can also determine whether a remote cache event of the other server is different from the local cache event and whether the local cache event is in conflict with at least one other cache event for the local data item. The processor can also determine whether the local cache event has a higher priority over the at least one other cache event and direct performance of the local cache event amongst the plurality of servers.