In the enterprise world a significant percent of applications (desktop and web) are written with the .NET framework. In the last decade, monitoring of .NET applications was focused on monitoring a) the health of server side components and b) samples of end-to-end transactions. Leave alone getting to the root cause in minutes, even discovering the entire app stack and all the transactions that flowed through it was cumbersome and not guaranteed.
With the shift of apps from desktop to web to mobile, shift of hosting platforms from the data center to the cloud, with the ability to analyze and correlate large sets of data cheaply and laser focus on scalability and User Experience - managing applications have transformed drastically in the last few years. Large scale enterprise .NET applications can now be diagnosed in a few clicks and in minutes. Here are some key changes in the way .NET application owners and operators are managing their applications today
Visual detection of bottlenecks through transaction-instance-level topological views: Today APM tools can discover your end to end distributed .NET application environment in minutes, showing you a topological view of all the components that your Common Language Runtime interacts with.
IBM Performance Management tools not only shows an aggregated transaction topology , but also shows the detailed topological mapping for single transaction instances , capturing network hops and sub-transaction nodes to help you see where the time is spent during that instance. Since 100% transactions are captured instead of sampling, you will not lose out on any key business transactions that may have failed.
Code level diagnostics: For every business transaction that fails or is slow, you can find out what line of code is causing the slowdown by looking at its stack trace. It shows the class name, method name and exact line of source code (e.g., SQL query to a MS SQL backend) that led to a slow request. Further you can see the pre and post code deployment patterns for your apps.
Applied log analytics: Searching for errors across the .NET application stack can be a laborious task. Earlier, while troubleshooting, operators and app owners would have to look through logs from different components independently. With integrated log analytics , IBM Performance Management lets you search for errors across log files for any component in your app stack in the context of the application. For example, you can correlate errors in your .NET framework’s CLR with an error in the MS SQL Server database that may be impacting a transaction.
One-glance health of all components in the app stack: As opposed to looking at multiple panes of glass to see details of your application’s health, today, at a glance in one UI you will be able to say the detailed health of all your app components. For example key metrics like time spent in Garbage collection, WCF status, CLR exception status , MS SQL Server space utilization and I/O etc.
Focus on End User Response Time: Today its not sufficient to understand how your app performs on the server side only. Apps are more and more heavier on client side code for quicker performance. IBM Performance Management shows exactly how long it takes to load a web page as experienced by the user.
IBM's Performance Management solutions are being used to manage some of the most complex and large scale .NET enterprise applications. In this quick video take a look at how you can drill down to the line of code in your .NET application with a few clicks
To learn more and do a free trial visit https://www.ibmserviceengage.com/application-monitoring/learn
Payal Chakravarty, Product Manager - IBM Application Performance Management