I recently presented at a large conference, and one of the slides I used showed the journey of a transaction through the IT infrastructure. Using it, I explain::(Windows, UNIX and Linux), many different network and storage devices, middleware all over the place and one or more mainframes. Companies found themselves using many different tools to monitor the health of transactions that touched all these various components, on many different platforms, monitoring and managing tens of thousands of transactions.Managing IT From a Business Perspective
We found ourselves managing the health of our transactions from a platform-technical perspective, creating silos in IT. We had a horizontal platform, database, network, and application silos and often vertical silos. Roughly around the year 2000, the business started to demand Service Level Agreements (SLAs) on a business level. It was simply no longer enough to let people know that even though a platform transaction was slow, all the individual components were doing fine. The business wanted more information; they were looking for the “who, what, where, when and why” of the story. But mostly, they wanted to know the problem was fixed. This forced IT to look at things from a whole new perspective; figuring out that something was wrong was already hard enough, fixing it manually every time, over and over again, was simply impossible. We needed more automation.
Most data centers today have only scratched the surface when it comes to really implementing automation to make transactions more reliable and run as autonomous entities. But the business is forcing IT to pay more attention to this. When things go wrong, IT simply can’t rely on manual interventions anymore. With the business driving cross-enterprise automation, IT must understand that a good set of point products is no longer sufficient if they aren’t built with cross-platform automation disciplines in mind.