公开(公告)号：US10229028B2

公开(公告)日：2019-03-12

申请号：US14659525

申请日：2015-03-16

Applicant: CA, INC.

Inventor： Mark Jacob Addleman ,  Jay Martin Cohen ,  Thomas Michael Repede ,  Gregory Neil MacKinnon

IPC: G06F11/07 ,  G06F11/30 ,  G06F11/32 ,  G06F11/34

Abstract: A system for managing application performance performs a learning phase and a monitoring phase. One embodiment of the learning phase comprises monitoring performance of multiple components of a software system to create first monitored component data for the multiple components and automatically identifying correlation between the components and a performance metric based on the first monitored data. The monitoring phase comprises monitoring performance of the multiple components of the software system to create second monitored component data for the multiple components, using the identified correlation to predict the performance metric, calculating the actual performance metric based on the second monitored component data, and reporting a performance problem if the actual performance metric differs from the predicted performance metric by more than a threshold.

公开(公告)号：US20170063966A1

公开(公告)日：2017-03-02

申请号：US14840809

申请日：2015-08-31

Applicant: CA, Inc.

Inventor： Charles J. Kalish ,  Lubomil Slivka ,  Gregory Neil MacKinnon

IPC: H04L29/08 ,  G06F9/455 ,  H04L12/861

CPC classification number: H04L67/10 ,  G06F9/45516 ,  H04L49/9005

Abstract: A method and computer product which re-directs sequential data to an off-platform system without modification of the source of the sequential data. An operating system on a first processing system is caused to create a privileged subsystem in the operating system. The privileged subsystem acquires sequential data output by a data producing application directed to a named dataset. An instruction specifies that the sequential data from the data producing application to the named dataset be processed by the privileged subsystem. A data forwarder is created on the first processing system. The data forwarder establishes communication with an external processing system and acquires the sequential data from the privileged subsystem. The data forwarder transmits the sequential data to the external processing system based on the instruction, and the instruction defines the external processing system as a destination for the sequential data.

Abstract translation: 一种方法和计算机产品，其将顺序数据重新定向到非平台系统，而不修改顺序数据的来源。 导致第一处理系统上的操作系统在操作系统中创建特权子系统。 特权子系统通过指向数据集的数据生成应用程序获取顺序数据输出。 指令指定从数据生成应用程序到指定数据集的顺序数据由特权子系统处理。 在第一处理系统上创建数据转发器。 数据转发器建立与外部处理系统的通信，并从特权子系统获取顺序数据。 数据转发器基于指令将顺序数据发送到外部处理系统，并且该指令将外部处理系统定义为顺序数据的目的地。

公开(公告)号：US20160274990A1

公开(公告)日：2016-09-22

申请号：US14659525

申请日：2015-03-16

Applicant: CA, INC.

Inventor： Mark Jacob Addleman ,  Jay Martin Cohen ,  Thomas Michael Repede ,  Gregory Neil MacKinnon

IPC: G06F11/34 ,  G06F11/30

CPC classification number: G06F11/3409 ,  G06F11/079 ,  G06F11/3003 ,  G06F11/3072 ,  G06F11/32 ,  G06F11/3428 ,  G06F11/3452 ,  G06F11/3466 ,  G06F2201/81 ,  G06F2201/865

Abstract: A system for managing application performance performs a learning phase and a monitoring phase. One embodiment of the learning phase comprises monitoring performance of multiple components of a software system to create first monitored component data for the multiple components and automatically identifying correlation between the components and a performance metric based on the first monitored data. The monitoring phase comprises monitoring performance of the multiple components of the software system to create second monitored component data for the multiple components, using the identified correlation to predict the performance metric, calculating the actual performance metric based on the second monitored component data, and reporting a performance problem if the actual performance metric differs from the predicted performance metric by more than a threshold.

Abstract translation: 用于管理应用程序性能的系统执行学习阶段和监视阶段。 学习阶段的一个实施例包括监视软件系统的多个组件的性能以创建多个组件的第一监视组件数据，并且基于第一监视数据自动识别组件之间的相关性和性能度量。 监测阶段包括监测软件系统的多个组件的性能，以便为多个组件创建第二监视组件数据，使用所识别的相关性来预测性能度量，基于第二监视组件数据计算实际性能度量，以及报告 性能问题，如果实际性能指标与预测的性能度量不同超过阈值。