Proactive Change Impact Management

Just my opinion here, but it’s probably best before releasing a change to a production system, for IT to have some idea of how that change will impact production. I’m guessing that your IT Release Manager would agree. However, when our research shows that 25% of all changes put into production cause some type of negative impact, and 10% of all changes have to be rolled back because they can’t be fixed…sort of indicates that something is missing from the change and release management process.

To be fair to the IT testing community, there are a several significant challenges to adequately test changes targeted for today’s multi-tiered software infrastructures. We’ve talked about these in the context of the IT Operations Perfect Storm, in which IT is pressed to test a high volume of different changes (everything from a patch to an application migration) against very complex software infrastructures that are interdependent with other systems (some not under IT’s direct control) in a business climate that expects nearly 100% uptime. So, some of what is missing is the commitment to testing fundamentals of budget, resources, expertise, and time.

Even for those organizations that make the preceding fundamental commitment, testing success can be limited by the time needed to build and maintain not just a representative staging and testing environment, but the tests, reports, and certification functions to facilitate testing. So part of the problem can be linked to a lack of comprehensive pre-deployment testing solutions.

Jasmine Noel of Ptak Noel and Associates, LLC participated with StackSafe in a webinar on Wednesday October 29th to discuss the other factor that is missing: a focus on effective process identified as change impact management. Change impact management (CIM) has commonly been defined as a method to evaluate change at a business service level to determine impact to the execution of a business process. Jasmine drove this definition a step further to the actual testing level.

CIM satisfies with three primary testing objectives:

• Validate that the configuration of a proposed change maps to established policy
• Thoroughly test changes to clearly identify the impact on production
• Monitor the behavior of the environment after the change is deployed

An effective CIM solution will:

• Draw conclusions based on real data
• Help IT understand complex environments
• Help IT understand the sequence of events to satisfy a change requirement
• Help IT understand how the system behaves based on a change
• Be easy to use to help IT focus on the results rather than process

Our experience shows that there are five additional factors that will ensure CIM success:

• Establish a formal staging and testing environment
• Test every change that is planned for production
• Test changes across the end-to-end infrastructure
• Utilize automated change management solutions
• Adhere to best practice guidelines

When you combine the commitment to test with the right solution and the right process, you should (as Jasmine states) “get the benefit of [testing] hindsight without actually living through a full-scale disaster!”

Popularity: 5% [?]

Filed Under: Change Management, IT Operations Research, Testing