An apparatus which provides the user with an indication of the computer network's health. The indication is provided by a network management station on the computer network. The network management station has a distributable piece of code which instructs agents to gather diagnostic and status information. The network management station then evaluates the network specific diagnostic and status data gathered by the agents. Based on the evaluation, the network management station generates a representation of the computer network's functionality (i.e., its "health"). Thereby, the user can readily determine whether the computer network requires repairs.

Method and apparatus for determining the health of a network

The invention comprises a method and system for monitoring the performance of a computer coupled to a computer network and, optionally, generating alerts when the performance of the computer has reached an alertable level. The method of monitoring the performance of the computer comprises repeatedly obtaining performance data using the computer wherein at least one performance value comprises a measure of the performance of the computer. The performance data is automatically sent over the computer network to a second computer coupled to the computer network. The performance data is then logged to a performance database using the second computer.

Method and system for performance monitoring in computer networks

An apparatus for monitoring the performance of a computer system. A number of performance monitoring hardware elements may be placed throughout a computer system to simultaneously monitor the performance of a number of distinct components within the computer system. An advantage of the present invention over a software based approach is that the present invention allows any node within the computer system to be monitored. In addition, the present invention does not run on the systems CPU and therefore the performance monitoring function does not decrease system performance while operating. Finally, because the present invention does not run on the system's CPU, the results of the performance monitoring function may be more accurate than a software base approach.

Performance monitor for digital computer system

A method and system within a data processing system are disclosed for halting execution of instructions by a processor in response to an enumerated occurrence of a selected combination of internal states within the processor. The processor includes a number of state machines and a means for monitoring the states of the number of state machines. According to the present invention, a selected combination of states of a subset of the state machines is specified. An enumerated occurrence of the selected combination of states of the subset of the state machines is then detected. In response to the enumerated occurrence of the selected combination of states, execution of instructions by the processor is halted such that states of the number of state machines within the processor remain substantially unchanged following the enumerated occurrence of the selected combination of states. In a first embodiment of the present invention, the selected combination of states is specified as a combination of states corresponding to an access to a specified address within an associated memory. According to a second embodiment of the present invention, the selected combination of states is specified as a selected sequence of states of a subset of the number of state machines.

Method and system for halting processor execution in response to an enumerated occurrence of a selected combination of internal states

Program activity graphs can be constructed from time-stamped traces of appropriate execution events. Information about the activities on the k longest execution paths is useful in the analysis of parallel program performance. In this paper, four algorithms for finding the near-critical paths of program activity graphs are presented and compared, including an efficient new algorithm that utilizes slack values calculated by the critical path method to perform a best-first search in linear space. The worst-case time and memory requirements of the new algorithm are in O(ke) and O(k+e), where e is the number of edges in the graph. Results confirming the efficiency of the algorithm are presented for five application programs. A framework for utilizing the near-critical path information is also described. The framework includes both statistical summaries and visualization capabilities. >

Near-critical path analysis of program activity graphs

A hybrid performance monitor developed for MSPARC, a mesh-connected, message-passing multicomputer, is described. The development of the hybrid performance monitor was a cross-disciplinary enterprise requiring custom hardware and a range of software support including monitor code, driver interfaces, probe history acquisition and processing, graphical display, and application probe injection. Programmable hardware was designed to unobtrusively collect events on each node and maintain their accurate chronological order. This distributed collection system was coupled by its independent network to a central monitor where data selection and presentation techniques played an important role in the visualization of the parallel system's execution. >

https://ieeexplore.ieee.org/iel2/617/5226/00202424.pdf

D. Linder

Papers

A performance monitor for the MSPARC multicomputer