Andy Wild is the president of OpTier.
Traditional approaches to Application Performance Management (APM) are usually based upon tools that act as individual components of your application stacks. This approach is flawed because it gives you a one-perspective view of the application. It shows the application server or database, but might neglect the end user, process or transaction view.
A new era of tech events has begun
We’re back in New York this November for the 4th edition of our growth-focused technology event.
Next-gen APM solutions take a broader view, with attention paid to users, performance, and the system as a whole. The benefit for businesses is greater context into transactions and the tools needed to analyze and manage applications properly. A cloud-based APM driven by analytics can strengthen your strategy
1. Detect and isolate problems
Multiple tools mean it’s hard to find the exact problems and detail the business and user impacts. With hundreds of server crash or failed disk alerts, tying them to quantified issues is very challenging. And you have redundancy in place, so failures might not affect the app SLA.
You need end-to-end transaction tracing to show how long applications take to execute and the time they spent in each tier, whether it’s LDAP, middleware, database, or Web server. This is results-oriented APM, where you focus on the affected tier that is impacting users. What’s the payoff? Greatly improved efficiency in finding root causes of problems.
2. Communicate across groups
Does your IT department engage in the “blame game?” The siloed groups with their own monitoring tools always insist their area isn’t the root cause of an issue. IT needs to understand their specific tool isn’t providing the broad view needed by the business.
Consider SQL response times. Database response times might be three seconds, but if the monitoring doesn’t catch the “Checkout” transactions take a minute, then you’re missing details.
You need a horizontal monitoring platform that includes all tiers. Transaction performance should be segmented into the logic of your application, so you can instantly grab the context of the problem, which is essential for judging the severity and fix for any issue. Support at all tiers means one team can respond to the fix instead of pandemonium and finger pointing.
3. Find the root
End-to-end transaction monitoring means you find the transaction and tier causing problems and get closer to the root cause. Isolation of an application or database tier means you get answers quickly, and your solution should also offer insight to the code and SQL levels for even greater context.
4. Truly end-to-end visibility
With cloud-based solutions, you need to offer more than just reports of five nines availability on CPU utilization or server availability. Every stakeholder can login to the system to see how long it takes to log in, search, or use various areas of the site/application. End-to-end tracing means you can report on speed, volume and SLA based on business transactions.
IT is under constant pressure to be faster and more reliable. Management no longer accepts multimillion projects with three year completion timeframes. Cloud-based APM solutions feature simple installation and are cost-efficient ways to get IT and the business focused on driving growth instead of monitoring infrastructure.
Despite tech advances, modern IT continues to grow in complexity. APM solutions can combat this complex nature by being more strategic and analytics to give users visibility into root causes and end-user impacts.
Cloud is the next step for APM, because the platforms are easy to use, dig deeply into issues, and promote improved operational efficiency and productivity.