Historically, in their efforts to monitor different infrastructure platforms, IT organizations have implemented a variety of fault and performance monitoring tools—many of which overlap in capabilities and functionality. While these tools excel at providing visibility into their specific platforms, they have the effect of creating IT silos because they lack the ability to provide an integrated view of the IT environment as a whole.
That’s problem number one. Problem number two comes into play when trying to break down the infrastructure silos created by these monitoring tools.
To break down silos, many IT organizations implement a “Manager of Managers” (MOM) approach; essentially they use event correlation tools to monitor their infrastructure monitoring tools. Thisdoes,provide the organization with a level of enterprise monitoring, but it’s still based on a bottoms-up approach.
Even though support teams can work together and analytics can be applied to data from the MOM, the monitoring is still dependent on the correlation and integration of data from each individual infrastructure platform. As a result, organizations often can’t see the forest for the trees. They understand what’s happening at the most basic level because they can see anomalies or look backwards to research the root cause of an incident, but they don’t have a real picture of the overall health of their IT.
Part of the reason this has happened is because modern businesses have evolved over the years to become increasingly dependent upon a portfolio of applications and services to generate revenue. The business is concerned with how these applications are performing for the end user, so it makes sense that they would want to monitor the health of these applications. The trouble is, again, that looking at the health of each individual app—or even at the health of a group of apps—tells you more about the application or group of applications than it does about the health of your broader IT.
As we see it, the future of enterprise monitoring is different. It should switch to a top-down approach which focuses on the need to understand the end-user experience. So: how can an organization monitor the overall health of its IT, while also keeping an eye on the health of its individual applications and the experience of its end users?
A key to executing this top-down approach is leveraging application performance monitoring (APM) and advanced analytics capabilities. APM tools can capture data about end-to-end latency, execution correctness, and quality, while also utilizing synthetic transactions to provide a true picture of performance for the real end users of the application.
APM tools also provide application topology, business transaction discovery, user-defined transaction profiling, and application-component deep-dive capabilities. With APM tools, it’s possible to not only applyadvanced analyticsto traditional data, but also to business data collected from transactional payloads. What’s more, APM tools move monitoring from being IT-centric toservice-centric.
Moving to this top-down approach to enterprise monitoring with a focus on end-user experience will allow your organization to quickly survey the application landscape (i.e., see the forest). At the same time, it will allow you to spot trends and real-time danger signs via dashboards and visualizations for any number of different roles–operators, administrators, development and engineering, QM, business users, etc. Once an issue is identified it can be quickly triaged and routed to the appropriate support group, who can then use the platform-specific tools to view system-level data and metrics to resolve incidents or research the root cause of a problem.