Browse Prior Art Database

Dynamic real time event prioritization

IP.com Disclosure Number: IPCOM000185633D
Original Publication Date: 2009-Jul-29
Included in the Prior Art Database: 2009-Jul-29
Document File: 2 page(s) / 45K

Publishing Venue

IBM

Abstract

This disclosure outlines a load balancing technique that can be implemented in Netcool Omnibus probes that would allow for the dynamic prioritisation of high priority events. This technology would prove to be most beneficial under event storm conditions where standard buffers are quick to be overloaded. Initial investigations suggest that the methodology could prove to significantly increase the throughput of high priority events in the case of an event storm.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 55% of the total text.

Page 1 of 2

Dynamic real time event prioritization

Currently when event storms occur in the Network Management arena, there is a high possibility that critical events could sit in a buffer for an unsatisfactory amount of time. This can mean there could be latencies of several hours before information about critical alarm conditions are made available to operators. The current solution is just to blindly buffer events and procedurally process them during an event storm. This solution is inadequate since critical events will be held in a queue until all previous events, critical or otherwise, have been processed. Since the existing solution allows for an unacceptable latency with regards to what could be high severity situations, it may cause serious financial ramifications for the affected business.Our proposed solutions aim to minimise the possible effects that an event storm could have on a business, both operational and financial.

    The core idea is to prioritise events being processed by Netcool Probes via a priority based algorithm to allow for higher severity events to gain events over those that are less important. The prioritisation can be controlled by rules files construct - so it is configurable and dynamic, and could allow for finer grained control for the end user.

    The advantage of this approach is allowing higher severity events to be prioritised thus meaning that business critical events can be seen far sooner then they otherwise would have been.

    The idea is to have a priority based queue in the probe. Rules file processing occurs before data is put onto the queue. During rules file processing it is possible to prioritise the event so that it can be processed before the other events. Essentially events can be marked as Low Priority(-1),Medium(0),High(1). The following rules file function is supplied to set the priority for an event.
setPriority(signed int priority)

    By default if the function is not used the priority will be assumed to be medium. High priority events get sent to the object server before medium or low priority events.

    The probe will have a property called PriorityWeightingFactor which will be used to determine how events get prioritised. It will have the following format

,

,

So for...