Setting up a tool that performs some type of end user performance testing is not enough, it is a type of testing that provides a view of the end user experience of using a part of a specific service. Adding Service Level Management on top of the testing is still not enough.
Business Service Management is a bit more encompassing, when there are slow response times, which piece of the supporting technology is the culprit, is this something that can (or needs to) be addressed now? If we were to take this slow database offline in order to address the issue, what impact would that have on the enterprise or end users. Business Service Management helps with these problems and more. End user response time measuring is just a piece of BSM, it might be a good starting point, but don’t be fooled, you are not done.
Remember there are several layers in the OSI model and having a health indicator from each of those layers (or several at least) is going to provide a better picture end to end of the health of the service. The big management tool vendors typically compete against each other, the typical model is rip and replace, they sell you new tools and get you to stop using the old tools… very expensive and disruptive proposition. Since there is no single vendor that is the best of breed for each of the OSI layers, then a single vendor for the end to end management doesn’t make sense. It makes more sense to purchase some of the tools to do specific types of monitoring, leverage opensource to monitor some of the other aspects and then roll all of these together into a single end to end view. This approach allows you the ability to swap out tools when they become dated or when the vendor is trying to hold you hostage at renewal time.
Having a single console that is able to integrate with all of the underlying technologies managing the environment and providing an end to end view is a better way to manage the enterprise, using a response time tool and crossing your fingers that everything will work out is risky.
Tobin
