Redgate Software launched its fourth annual State of Database DevOps Survey.
In an age of big data, more must mean better, right? Scan the APM marketplace and you’d think that there are legions of data nerds eager to swim in data, painstakingly applying their tightly guarded expertise to reduce their MTTI (mean time to innocence). The reality is that no one in IT has the time. It’s answers, not data that’s paramount.
How did we get to this state of data bloat? The biggest cause is not fully understanding the real-world users of application monitoring. Just as IT managers know that great APM starts with the end user experience, vendor product design must start with those IT users in mind. Sounds obvious, yes, but it’s too easy to fall into a vicious circle of capturing and reporting every metric with the assumption a human is eager to own the forensics.
As discussed in a previous blog, selecting the right APM for DevOps is an “EPIC” decision. Easy, Proactive, Intelligent, and Collaborative is a user-driven approach to APM focused squarely on helping ITOps teams succeed at managing application performance.
To build an E.P.I.C. APM solution, requires not only a clinical understanding of the most common needs by role type, but also working with real users; understanding a day in their lives and truly empathizing with their challenges. How do you make the clickpath to answers truly intuitive? What’s the preferred way to easily share insights with colleagues? Answers to these and other questions means really spending a day in the life with many different roles such as application developer, APM administrator, level 1 support analyst, production support analyst, middleware specialists, test engineers, production performance engineers and more.
The developers of an E.P.I.C. APM solution must understand the stories behind each of the roles and at a very practical level, what steps and methods they enact for success. And here’s the kicker: to create an APM solution that its users are passionate about, you must be passionate about them. So, in addition to having deep day in the life insights, the developers of an E.P.I.C. APM solution must truly empathize with and care about the success of each role.
Here are some examples of how E.P.I.C. APM delivers on its passion for APM users:
Making APM easy to adopt, fast time-to-value, simple-to-use, easy to manage and configure. For example, simplify management for thousands of agents with a central repository of all agent configurations and meta-data across all of your APM clusters in minutes and not hours.
In agile environments the concept of “canary-testing” new code (against unsuspecting) users has grown in popularity to gain early detection and prevent bigger issues during the full push of a revision. Proactive approaches to APM ensure that insights are gathered quickly and shared with development, versus the ‘wait and see’ tactic.
Collecting and delivering data is one thing, making it actionable is another. APM can be smarter. For example, automatically detecting degradation in one user’s experience and pinpointing the code or even infrastructure that is the cause. Instead of manually digging to get call stack visibility, transaction traces for that particular issue are surfaced automatically.
Enable better communication between Dev and Ops specialists to resolve problems faster by utilizing the same production tool in development, and with a unified view of the infrastructure and apps that affect business services. For example, operations providing real world data to development to make enhancements to apps more relevant and improve performance.
E.P.I.C. may sound like a clever marketing acronym but when we speak with the real front-line users of APM these are the areas that they care the most about. Designing for their unique needs is producing a new number of role-specific features that help convert big data into big answers.
Kieran Taylor is Sr Director, Product & Solutions Marketing, APM & DevOps, CA Technologies .