发明名称 Using correlated stack traces to determine faults in client/server software
摘要 A method, apparatus and computer program product for using correlated stack traces to determine faults in client/server software is presented. Client/server software is executed. A client call stack is maintained on the client and a server call stack is maintained on the server. When an unexpected occurrence is detected while running the client/server software then the client call stack is collected, as is the server call stack. The client call stack and the server call stack are correlated to produce a correlated call stack and the correlated call stack is provided to a user.
申请公布号 US9064046(B1) 申请公布日期 2015.06.23
申请号 US200611325285 申请日期 2006.01.04
申请人 EMC Corporation 发明人 Cormier Richard Francis;Michel Ruben
分类号 G06F9/44;G06F11/00;G06F11/36;G06F11/32;G06F11/30 主分类号 G06F9/44
代理机构 Krishnendu Gupta 代理人 Krishnendu Gupta ;Kazanjian Gerald P.
主权项 1. A method comprising: executing, via at least one processor device, client/server software application code in a distributed system; maintaining a client call stack on a client comprising at least one client stack frame pushed onto the client call stack by operation of methods of client/server software executing on respective client threads; maintaining a server call stack on a server comprising at least one server stack frame pushed onto the server call stack by operation of methods of client/server software executing on respective server threads; detecting an unexpected occurrence in the execution of the client/server software application code executing on the at least one processor device in the distributed system; collecting the client call stack from the client, collecting the server call stack from the server; correlating the client call stack with the server call stack to produce a correlated call stack comprising a first sequence of frames from the client call stack, a second sequence of frames from the server call stack, and a stack frame indicating a partition between the first sequence of frames from the client call stack and the second sequence of frames from the server call stack; providing the correlated call stack to a user; and using the correlated call stack to determine a root cause of the unexpected occurrence detected in the execution of the client/server software application code in the distributed system.
地址 Hopkinton MA US