Abstract: This paper proposes and discusses a new type of tool for Linux based software systems analysis, testing and optimization as well as the new approach which is based on this tool and will help to define the best moment for executing of effective and smart software upgrade in automatic mode on active and online system. The best moment means one when the software upgrade will cause the minimal services losses. The presented tool is called CAP (Characterization of Availability and Performance) and provides engineers with an instrument for systems performance tuning, profiling and issues investigation as well as stress and load testing. The described system operates with over 150 Linux based system parameters to optimize over 50 performance characteristics. The paper discusses the CAP tool’s architecture, multi-parametric analysis algorithms, and application areas. Furthermore, the paper presents possible future work to improve the tool and extend it to cover additional system parameters and characteristics. The prediction of the software upgrade (SU) best moment mentioned above is supposed to be made on basis of performance and availability statistics got from the CAP tool.

Keywords: Software Systems Analyses, Linux Servers, Telecommunication System, Performance, Availability, Serviceability, Software Upgrade Prediction.

ACM Classification Keywords: C.4 Computer Systems Organization - Performance of Systems - Reliability, availability, and serviceability.

Link:

SOME APPROACHES FOR SOFTWARE SYSTEMS ANALYSES AND ITS UPGRADE PREDICTION

Igor Karelin

http://foibg.com/ijita/vol16/IJITA16-2-p05.pdf