This is a demo of the AppD / Turbonomic integration completed thus far.
We begin in the AppD interface, monitoring a single application, LexCorp’s Webmail, which is composed of a web tier, a middleware tier and a database.
We begin by adding load to the Tomcat server, seen here by response time spiking at around 1:50 PM. This load is causing significant performance issues across the entire application, as you can see in the Web and Middleware tiers, as indicated by the Red circles here in AppD.
Switching to Turbonomic, we see that Turbonomic has not only discovered the WebMail app, but also its entire topology including the Tomcat and Apache nodes we saw in AppD.
With this integration, Turbonomic takes the data from AppD, stitching and correlating it with the rest of the infrastructure beneath the VM.
We accomplished this by simply adding a Cisco UCS Manager target along with a vCenter target and a NetApp target to the environment, enabling us to understand the full topology from the Chassis to the switch and disk array, to the hosts themselves up into the VM.
Within Turbonomic, we take information from these targets and application data from AppDynamics like response time, application heap and other performance metrics to help the environment manage itself back to health. Here, application response time has increased significantly due to the stress being placed on the CPU in the VM hosting the Tomcat node.
Turbonomic automatically generates an action to scale the VM to increase VCPU capacity, which we take by ticking the box and clicking apply selected. The action is then executed by vSphere automatically.
Returning to AppD, we see the application response time returning to a healthier state after taking the scale vCPU action. Looking deeper at response time, we see that it is rapidly falling from its earlier peak due to this scaling action. Looking back further, we see the direct correlation between taking action and improving response time.
In Turbonomic we see the VM is now back in a healthy state, returning to green from yellow, due to our scaling action increasing CPU resources, helping satisfy the increased load on the VM.
In AppD, we see the same – that everything has moved back to a healthy state.
Everything we did here manually can be automated in Turbonomic, enabling the environment to manage itself and assure performance.
This is the power of the AppD – Turbonomic integration. Connecting AppD’s understanding of an application’s logical flow with Turbonomic’s understanding of the real-time resources needed for the application to perform. Enabling the autonomic data center.