17550389. CONTENTION DETECTION AND CAUSE DETERMINATION simplified abstract (INTERNATIONAL BUSINESS MACHINES CORPORATION)

From WikiPatents
Jump to navigation Jump to search

CONTENTION DETECTION AND CAUSE DETERMINATION

Organization Name

INTERNATIONAL BUSINESS MACHINES CORPORATION

Inventor(s)

Murtaza Eren Akbiyik of Zürich (CH)

Anastasiia Didkovska of Boeblingen (DE)

Dorian Czichotzki of Stuttgart (DE)

Dieter Wellerdiek of Ammerbuch (DE)

CONTENTION DETECTION AND CAUSE DETERMINATION - A simplified explanation of the abstract

This abstract first appeared for US patent application 17550389 titled 'CONTENTION DETECTION AND CAUSE DETERMINATION

Simplified Explanation

The patent application describes a computer-based method for identifying the cause of a performance issue in a computer system that is running different workloads. Here are the key points:

  • The method involves receiving system performance data from the computer system.
  • The received data is then separated into two categories: contention-related data and non-contention related data.
  • The contention-related data is further divided into two parts.
  • The first part is inputted into a machine-learning system with a trained model to predict instances of contention and their impact.
  • The second part is scaled with the predicted impact values and fed into a second machine-learning system with another trained model.
  • The second machine-learning system predicts contention instances and their impact values for the different workload groups.

Potential applications of this technology:

  • This method can be used in various computer systems, such as servers or cloud computing platforms, to identify and address performance anomalies.
  • It can help system administrators or IT professionals troubleshoot and optimize the performance of complex computer systems.

Problems solved by this technology:

  • Performance anomalies in computer systems can be difficult to diagnose and resolve, especially when multiple workloads are running simultaneously.
  • This method provides a systematic approach to identify the cause of performance issues by analyzing contention-related data and predicting their impact on different workload groups.

Benefits of this technology:

  • By accurately identifying the cause of performance anomalies, system administrators can take targeted actions to improve system performance and minimize downtime.
  • The use of machine learning models allows for automated and efficient analysis of system performance data, reducing the time and effort required for troubleshooting.


Original Abstract Submitted

A computer-implemented method for identifying a cause of a performance anomaly of a computer system executing workloads in different workload groups is disclosed. The method comprises receiving system performance data, separating contention-related data and non-contention related data within the received system management data, feeding a first part of the contention-related data to a first machine-learning system comprising a trained first machine-learning model for predicting first contention instances and related first impact values as output, and feeding a second part of the contention-related data scaled with the first impact values to a second trained machine-learning system comprising a trained second machine-learning model for predicting second contention instances and related second impact values for the different workload groups as output.

Retrieved from "http://wikipatents.org/index.php?title=17550389._CONTENTION_DETECTION_AND_CAUSE_DETERMINATION_simplified_abstract_(INTERNATIONAL_BUSINESS_MACHINES_CORPORATION)&oldid=26269"