ITIL Service Modelling, or Service Impact Modelling, has been a buzz-word for a long time. It’s a great concept and the benefits to the business are huge. But while many companies have tried to implement service modelling, most have failed. I’ve seen service modelling projects fail over and over again due to inadequate tools, manual processes, and ever-changing IT environments. Still, despite the failures, the lure of the service modelling sirens is hard to resist and projects will continue to sail towards the sirens, with little or no success.
What are Service Impact Models really?
A service impact model is a model of the infrastructure and application layer supporting a service. When an event occurs, the model is used to calculate the impact of the event on the service availability. If the model is not 100 percent correct at all times, it is virtually useless – this has been the downfall of the service model.
The Paradigm Shift
To succeed with Service Impact Modelling we need to be clear about one thing: The model only models the reality. Think about that. What everyone’s been trying to do is to have the model match the reality 100 percent, including servers, relationships, routers, and maybe even network cards. As the reality changes, the model needs to be updated and this has proved too complex and difficult.
What’s the solution?
Simplify. (Remember: The model only models the reality.) Then simplify again.
Removing the modelling aspect removes most of the complexity of the problem. The concept is simple:
- Create a service or application
- Relate all supporting entities directly to the service or application
- Hook in real user monitoring to accurately measure the impact of events on the end users
Take the example of a simple web application consisting of three load-balanced web servers and a mirrored database backend. If we relate all servers, load balancer, application containers, etc. directly to the service, we lose the ability to accurately calculate the impact of events to the service. However, by measuring each and every end-user transaction, we get an accurate picture of how events are impacting the end user. IT can remove the complex modelling process, replace it with an automated tool, and lose nothing.
This is precisely what BMC Easy Service Impact Analysis accomplishes. You get the full benefits of Service Monitoring, including event prioritization and intelligent root cause analysis, without the complexities of Service Modelling. I’ve worked with several customers who have already automated and implemented Easy Service Impact Analysis with great success – one of them with several hundred services!
While the Achilles’ heel is still the relationships created in step two, the model is so simple that this is easy to automate. It can even be as simple as importing an Excel sheet every time it’s updated with new servers, or applying a small set of rules to an automated Discovery scan.
For more information about BMC Easy Service Impact Analysis, contact me at firstname.lastname@example.org. You can also learn more about these unique solutions that enable and automate Easy Service Impact Analysis:
These postings are my own and do not necessarily represent BMC's position, strategies, or opinion.