April 5, 2016 | Written by: Ben Stern
Share this post:
Monitoring complex business applications has always been a difficult task. As more applications move to the cloud and include dynamic provisioning of resources, the task becomes even more challenging.
Companies would like to identify precursors to an application outage or slowdown within performance characteristics.. With many servers, network switches, storage and hypervisors supporting the business application, this can be a daunting task.
Analyze and identify the problem
IBM application performance management (APM) does an outstanding job of monitoring business applications and helping to determine root cause. But there are types of problems where APM doesn’t provide enough visibility and proper early detection. For example, within a small enterprise there is typically more than a terabyte of performance and log data processed every day. Additional capabilities are needed.
Fortunately, IBM provides IBM IT operations analytics: a family of products designed to analyze these massive quantities of data. First, take a look at the log data. All components like network switches, storage and operating systems have log files with useful data. In addition, all of the applications and middleware that make up the business applications log useful information, with each storing data in different formats.
IBM Operations Analytics – Log Analysis
IBM Operations Analytics – Log Analysis can centralize all of the log data, store it in a useful format and analyze it. Patterns of problems that might be affecting your business applications can be seen and monitored. Pattern detection and search, along with expert advice, helps customers diagnose and solve problems quickly.
This is invaluable in reducing mean time-to-repair. And that means cost savings.
IBM Operations Analytics – Predictive Insights
A second application of cognitive analytics is the use of streaming analytics to analyze performance metrics in real time. IBM Operations Analytics – Predictive Insights is a product that provides this capability.
Predictive Insights allows organizations to stream massive amounts of data to discover patterns and detect anomalies. After analyzing the business application for a period of time, Predictive Insights can establish baselines for normal behavior.
In addition, it can automatically identify the relationships between metrics on different servers and establishes a set of normal business application behaviors.
For example, when transaction rates through an HTTP server increase, there may be a corresponding decrease in response time and an increase in KPIs on the messaging system, the J2EE server and the database server. Those patterns are learned, so when the behavior of one of the key application components starts to misbehave, customers can be alerted to the abnormal behavior.
In addition, the product can detect the early characteristics of a problem that will ultimately cause an outage. In some cases, these patterns can be detected hours or even days before the problem occurs. Take a look at a specific scenario.
cognitive APM application performance management
As seen above, the black line represents response time for the application. In the pink shaded area, an anomaly was detected and an alert was sent to the operations team.
This was a multi-variant anomaly.
IBM IT Operations Analytics – Predictive Insights automatically determined that two database metrics were related to the response-time problem being observed. As the operations team debugs the problem, having this information is critical. Instead of debugging the application server, they can focus on the database, which started behaving abnormally first and affected the application server in turn.
As you can see, analytics can be a powerful tool in your arsenal to keep your business applications up and running smoothly. Combined with APM, you can detect problems early, isolate problems and keep your business applications running smoothly. Learn more about IBM IT operations analytics products here.
Explore the APM smartpaper or register for the Gartner webcast discussing the future of APM.