Application Performance Management (APM) have existed in their modern form since 1996, when Wily founded the business of monitoring custom developed Java applications in production. Today, many people still think of APM as just a solution for these custom developed applications.
It is true that Agile Development and DevOps have fueled the need for more monitoring by APM solutions of custom developed applications, but these solutions only focus upon one part of the problem – issues with the code in production.
Traditional developer focused APM tools (let’s call these DevOps focused tools), focus just upon the code and upon helping the developer supporting custom code in production find issues with that code.
There is a requirement for a new and different set of APM solutions that is not limited to just code problems and developers as the audience. This requirement is for APM solutions to work for every application and that help the Operations team support every application in production.
This new category of APM solutions is called the Application Operations or AppOps category of APM solutions, and it focused upon the Operations staff who is responsible for supporting every application (physical, virtual or cloud), running across every operating system (Windows or Linux), and every source of application (purchased or custom developed). The key attributes of an AppOps focused APM solution are:
1. Real-Time and Comprehensive Data Collection
Any tool that your Operations team uses needs to be able to monitor and capture data from every application, all the time. Without that, you’re likely going to be stitching multiple tools together, prohibiting your team from understanding how their applications perform in context with each other.
2. Continuous, End-to-End Measurement of Application Response Time
Operations teams in the virtualized world are now responsible for multiple applications running simultaneously on a shared set of infrastructure. Because of this newfound and complex responsibility, Operations now needs a tool that measures an applications response time across every applications system, the key metric that both IT and business can relate to( no matter how the app was developed).
3. Dynamic Discovery of Applications and their Topology
Time is money, as they say. Because of how often new applications are added into virtualized ecosystems, it’s necessary that Operations has a tool that auto-discovers new applications, names them, and maps their topology without a team member lifting a finger.
4. Deterministic Root Cause Analysis (opposed to inferred root cause)
Deterministic Root Cause is the only method that works for both resource utilization based issues and issues caused by changes in configuration, promising Operations teams immediate answers to the question of what issue, in the infrastructure, is causing the application performance issue.
5. Broad Application Platform Support
Code-level monitoring tools can only monitor custom-built apps, which represent 20% of all apps. AppFirst provides insight into both custom apps and packaged apps, which represent 80% of the application market.
6. Zero Initial and Ongoing Configuration
Modern cloud and virtualized environments are simply too dynamic and potentially distributed for approaches that require heavy configuration and customization to work. Therefore, it’s essential that Operations teams have a solution that works out-of-the-box with zero upfront and ongoing configuration as apps are changed and added.