2020
2017
System and method for determining relevance of application software maintenance
Kulvir S. Bhogal, Samir A. Nasser
software, operating system, data store, computer science, computer program, application software
Abstract
A method, computer program product, and computer system for compiling, by a computing device, a list of hosting software classes included in the hosting software fix pack when a fix is available. An execution path of each application hosted on a hosting software may be recorded. The execution path may be stored in a data store for each application. It may be determined which operations of each application interact with the hosting software. The operations of each application used at runtime that interact with the hosting software may be stored, including storing invoked hosting software operations and classes used by the operations of each application. The invoked hosting software operations and classes used by the operations of each application may be compared with corresponding operations and classes provided in the list included in the hosting software fix pack. A list of each intersection of the comparison for each application impacted by the hosting software fix pack may be generated.
2015
SYSTEM AND METHOD FOR PREDICITING RESOURCE BOTTLENECKS FOR AN INFORMATION TECHNOLOGY SYSTEM PROCESSING MIXED WORKLOADS
Samir A. Nasser
real time computing, performance metric, information technology, information processor, hot spot, computer science, computer network, bottleneck
Abstract
A method of predicting resource bottlenecks for an information technology system processing mixed workloads includes determining, through a processor, a probability that one or more of a plurality of requests will access one of a plurality of resources, determining, through the processor, a performance metric of the one of the plurality of requests, identifying, through the processor, a potential hot spot based on the performance metric, calculating, through the processor, a probability that each of the plurality of requests will concurrently execute on the one of the plurality of resources, and providing, through the processor, an alert predicting that a bottleneck could occur at the one of the plurality of resources.
SOFTWARE REQUEST-FILTERING PREDICTIVE TECHNIQUE BASED ON RESOURCE USAGE PROBABILITIES
Kyle G. Brown, Samir A. Nasser
native image generator, runtime callable wrapper, runtime verification, entry point, tree traversal, filter, software, parallel computing, real time computing, computer science
Abstract
In filtering requests to be forwarded to a runtime environment, a filtering apparatus intercepts a new runtime request for the runtime environment and determines execution paths that may be traversed by the runtime request when executed in the runtime environment. The filtering apparatus assigns a probability of traversal by the runtime request to each of the execution paths and identifies at least one given execution path that reference a stressed resource of the runtime environment. Based on the probabilities assigned to the at least one given execution path, the filtering apparatus determines whether or not to block the runtime request from being sent to the runtime environment. If the probability assigned to the at least one given execution path exceeds a configured threshold, the runtime request is blocked from being sent to the runtime environment. Otherwise, the runtime request is sent to the runtime environment.
DYNAMIC SYSTEM SEGMENTATION FOR SERVICE LEVEL AGREEMENTS ENFORCEMENT
Peter C. Bahrs, Kyle G. Brown, Samir A. Nasser
service level agreement, change request, enforcement, service level, risk assessment, real time computing, computer science
Abstract
In service level agreement enforcement, a segmentation module intercepts a request for a current runtime environment, extracts request data from the request, and passes request data to a risk assessment component. The risk assessment component finds a request history profile matching the request data, determines a request impact on runtime environment resources from the request history profile, and obtains the current state of the current runtime environment. The risk assessment module determines the risk to the service level agreement associated with the request or any other request currently executing if the request is serviced in the current runtime environment, based on the request impact and the current state of the current runtime environment, and outputs the risk assessment to the segmentation module. When the risk assessment indicates a risk, the segmentation module passes the request to an isolated runtime environment. Otherwise, the request is passed to the current runtime environment.
AVOIDING WEB REQUEST FAILURES BEFORE THEY OCCUR BY COMPONENT ANALYSIS
Kyle G. Brown, Samir A. Nasser
change request, component analysis, probabilistic logic, web service, distributed computing, computer science
Abstract
Disclosed herein is a method for predicting and avoiding request failures. The method includes receiving a request for access to at least one web service and analyzing the request to identify at least one probabilistic resource tree for handling the request. The method further includes detecting a problem preventing a usage of a resource in the at least one probabilistic resource tree. The problem causes the request to fail upon implementation of the request. Furthermore, the method includes performing an action to avoid the implementation of the request.
