Traditionally, APM tools have done a poor job of monitoring and correlating application infrastructure performance with top-line application performance. This creates additional challenges when trying to resolve application slow-downs, because understanding if infrastructure resource contention has played any role in an application issue is critical. Don’t worry, Instana’s Application Perspectives makes it fast and easy to see if the application infrastructure is contributing to the latency of your application services.
Correlating Performance Information Between Applications and Application Infrastructure
Instana has always provided the best possible application infrastructure discovery and correlation. Once installed, an Instana agent looks to see what components are running, then automatically attaches a monitoring sensor to each component and then starts monitoring. One of the first results is a technology stack graph similar to this…
The image above shows that Instana automatically discovered and monitored (from right to left) a JVM, running inside a Java process, running on a docker container, running on a Linux host. This is the application infrastructure portion of the full technology stack (stored in the Instana Dynamic Graph), but how does Instana connect this stack to any running applications and services? We start by automatically discovering and monitoring the service running within the processes or run-time environments.
The image above shows that Instana has automatically discovered, monitored, and connected the Shipping service to the JVM in the Technology Graph above. As requests flow through the shipping service, EVERY request is traced and correlated with the appropriate services and infrastructure.
Application Perspectives and Infrastructure Performance Analysis
Now that you have a basic understanding of Instana dependency graphs, it’s easy to demonstrate how Application Perspectives works in perfect harmony with the existing infrastructure correlation.
Application Perspectives focus the analysis and display of your monitoring data on what YOU really care about. After you have created an application perspective you can easily find all of the infrastructure that provides the services that are members of your Application Perspective.
Clicking on the container name that has the largest latency we are immediately presented with an infrastructure dashboard for that container.
This dashboard screenshot above shows some minor throttling issues within the container itself but no other infrastructure issues stand out after exploring all of the dependant infrastructure. In this case we can easily rule out infrastructure contention as a cause of latency. It’s also important to note that the infrastructure metrics you see in the charts are 1 second granularity so that you never conceal an issue by using averaging.
Looking at the traces of slow requests (in the screenshot above) to the payment service we can easily see that the payment service is slow due to external service requests to our payment gateway (this has been artificially induced for demonstration purposes).
Instana’s combination of full stack, automatic dependency correlation along with Application Perspectives makes it easy and fast to understand the impact YOUR infrastructure has on YOUR running applications. Sign up for a free trial and check out Instana in YOUR environment today.