Browse Prior Art Database

Method and System to Manage Component Update Based on Feature Diagnostic Disclosure Number: IPCOM000238974D
Publication Date: 2014-Sep-29
Document File: 4 page(s) / 1M

Publishing Venue

The Prior Art Database


Analyzing the history of workload execution with the method described, it is possible to upgrade a workload scheduling solution in a safer way so that if there are some bugs in the new version to upgrade they will not completely block current workload execution

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 54% of the total text.

Page 01 of 4

Method and System to Manage Component Update Based on Feature Diagnostic

    In IT world existing scheduling software products have the goal to execute well defined workloads. These workload automation software are based on client-server architecture with agents (clients) connected to the server. Each agent is the executor of a workload planned on the server side. Often, these agents need a software upgrade to apply fixes or enable new features. In a large environment, with hundred or thousand of agents to upgrade, it may happen that many agents fail to upgrade or worse, the upgrade may succeed but the new version may have injected a new bug that will cause the next scheduling of workloads to fail.

    It is possible to accomplish the task to update a large number of scheduling agents splitting the whole number of agents in little groups and apply the upgrade to the first one of these subgroups, then verify if the new installed version has injected some bugs analyzing the history of successful/failed jobs on this group of agents and then, if a well defined successful criteria has been met, proceed with the next group of agents until all the agents of the environment have been upgraded. In case the defined successful criteria has not been met, the whole update process stops waiting for the analysis of an administrator.

    The scheduling system as described is able to track and maintain the history of all successful/failed jobs that run when the workflows are executed. A new...