Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Method to Prevent Malicious Events from Affecting System Behavior based on Policies

IP.com Disclosure Number: IPCOM000132027D
Original Publication Date: 2005-Nov-29
Included in the Prior Art Database: 2005-Nov-29
Document File: 2 page(s) / 24K

Publishing Venue

IBM

Abstract

Increasingly business systems and infrastructure are instrumented for events (business and IT), monitor those events, and the system behavior is changed based on those events. Using common event infrastructure and open standards help events to be the common mechanism to pass alerts, situations, data, etc between different systems and based on policy, they help the system to autonomically adjust to the situations. On the negative side, if such events are generated by untrusted parties, this would end up with false alarms and might take the system down the path of undesired behaviour and worst case, self-destruction. The core solution outlined in this article is to handle events based on policies that take into account the authenticity of the emitter, criticality of the event from the system viewpoint, suspicion factor based on context of event emitted. When taking actions, the systems also take into account the risk associated with not handling the event, cost involved when event is handled per 'steady state' policies (i.e assuming event is authentic, what happens when system reacts to the event, etc), and others.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Method to Prevent Malicious Events from Affecting System Behavior based on Policies

In an OnDemand Operating Environment using Common Event Infrastructure (CEI) as the event mechanism to help pass events between systems and services, system behavior is changed based on the events. On the positive side, this helps events to be the common mechanism to pass alerts, situations, data, etc between different autonomic managers and based on policy, they help the system to autonomically adjust to the situations. On the negative side, if such events are generated by untrusted parties, this would end up with false alarms and might take the system down the path of undesired behaviour and worst case, self-destruction.

Existing solutions in eventing mechanisms deal with events at the consumer side.. i.e, at the end point receiving the events and managing them. In simple systems where events are point to point and the identity of the producer is "trusted", this may help lessen the risk. But in the case of dynamic, on demand systems, where applications running on the infrastructure may vary, and that events may be routed along multiple intermediaries and managers, it may be too late to handle the filtering at the consumer side. What is required is to prevent the situation from happening, and preventing this closer to the producer itself. This article proposes an approach to address this problem.

Solution

Handling events at both the emitter side and the consumer side is necessary and combination of these will help prevent undesired behavior.

On the emitter side:

Ensuring authenticity of the event emitters is done close to the producer/emitter and this is based on various policies that can be specified. Such policies may include whether the application is trusted enough (e.g. where the code is loa...