Cisco AppDynamics Community

Claudia.Landivar · ‎01-17-2024

Intelligent monitoring in a rapidly changing digital landscape

Video Length: 4 min 17 seconds

CONTENTS | Introduction | Video | Transcript | Resources

In line with enterprise applications transitioning to the cloud, the need for simplified and AI-driven observability solutions is growing. AppDynamics’ cloud native response is an innovative platform designed to effortlessly onboard customer cloud environments, automate monitoring of ephemeral environments, and streamline MELT (Metrics, Events, Logs, and Traces) data correlation. By leveraging the power of data science—including machine learning and AI—it provides comprehensive solutions to challenges arising from applications, Kubernetes, infrastructure, or other cloud-native aspects in a multi-cloud world.

This demonstration video explores these capabilities, demonstrating AppDynamics’ effectiveness through an application deployed to an AWS Kubernetes cluster. See Helm charts used to simplify the installation of monitoring components in a Kubernetes environment and enable app deployment auto-instrumentation using OpenTelemetry.

Video Transcript

Spoiler

00:00:09:05 - 00:00:37:18
As applications move to the cloud with the need to simplify observability, we are meeting those needs with AppDynamics. Here, we will show an example of a simple application deployed to an AWS Kubernetes cluster. It contains one pod and two virtual machines to support the cluster.
We'll show how easy it is to onboard on to AppDynamics, requiring minimal configuration as is needed in large-scale environments and the world of microservices, as well as show how we are able to monitor every aspect of the application as it scales to meet user demand.

00:00:37:18 - 00:01:01:10
First, from the UI, we easily onboard a cloud connection to consume cloud components, including infrastructure metrics. Once that configuration is set, we can see how we pull in the infrastructure, including hosts, load balancers and storage, into the platform automatically. Reviewing the hosts, we see two hosts are being monitored through CloudWatch, including host metrics going back to investigate how many hosts are currently allocated to the Kubernetes cluster.

00:01:01:17 - 00:01:21:14
We see those (same) hosts, which we see in the AppDynamics UI, are the same hosts allocated to this cluster.
The cloud connection is a one-time setup. Any new hardware allocated in the cloud will be monitored automatically by AppDynamics, as we'll see later in the video.
After we set up our cloud connection, we will use Helm charts to simplify the installation of monitoring components in the Kubernetes environment.

00:01:21:16 - 00:01:41:13
This includes installing the app, mixed cluster operator and agent, the infrastructure and logs agents to monitor and collect logs.

Additionally, we will enable the ability to auto-instrument applications using Open Telemetry, which provides a framework for capturing application telemetry data.

To recap to this point, we have rapidly established a cloud connection to either a public or private cloud.
Additionally, with just a few steps, we have started gathering telemetry and logs related to the Kubernetes environment.

00:01:41:16 - 00:02:04:07
Next, we will auto-instrument the Pet Clinic application deployment using Open Telemetry. After playing some load to the application, the Pet Clinic service appears in AppDynamics, as shown in this flow map. We can also see the business transaction generated by this Pet Clinic service, which is coming from the Open Telemetry instrumentation.

00:02:04:10 - 00:02:21:17
Here, we see three business transactions are being generated from this service. We also see the service instance of which there is only one. We see the Kubernetes cluster, the services running under, which namespace, which workloads, and how many pods of which we know there is only one deployed, and which hosts the pods are running under, of which there is only one pod.

00:02:21:23 - 00:02:39:10
So, there is only one host out of the two that are running this pod. Remember, we didn't have to do any fancy configuration to enable all this.
We set up a cloud connection, installed a Helm chart to enable Kubernetes monitoring, and we auto instrumented our application. We didn't have to set up an application tier or node names as we do with a commercial SaaS solution.

00:02:39:10 - 00:03:01:05
And everything you see in AppDynamics was automatically ingested into the platform and automatically correlated between all entities, including APM, Kubernetes and the cloud infrastructure. Moving forward, AppDynamics will automatically monitor and show any changes to the infrastructure, such as scaling up or down, issues in Kubernetes or changes to the application, as well as any performance issues.

Let’s show this by scaling the application.

00:03:01:07 - 00:03:24:05
First, we see there is only one pod running for Pet Clinic service and two nodes are host to support the cluster. Now we'll scale up the Pet Clinic application service from one pod to 20 pods total. We can see all the parts started with one having an issue. So a total of 21 pods. We also see the Kubernetes cluster automatically scaled up the number of virtual hosts, EC2 instances in this case to handle the increased demand for 20 pods.

00:03:24:06 - 00:03:45:13
We now have 20 pods running, one pod failed, and five hosts for the infrastructure. Coming back to the AppDynamics UI and under our Kubernetes cluster, we see AppDynamics automatically monitored and correlated the five total hosts to this cluster and Pet Clinic service. Additionally, we're also now automatically monitoring and reporting on 21 pods to support the Pet Clinic service.

00:03:45:14 - 00:04:07:09
AppDynamics’ new Cloud Native solution will make it easier to onboard customer cloud environments and automatically monitor their ephemeral environments, and automatically correlate incoming MELT metrics, events, logs, and traces. Once the data is in the platform, we can apply data science, such as machine learning and AI, to help solve problems from the application, Kubernetes, infrastructure or other cloud native aspects that the customer is using.

00:00:09:05 - 00:00:37:18As applications move to the cloud with the need to simplify observability, we are meeting those needs with AppDynamics. Here, we will show an example of a simple application deployed to an AWS Kubernetes cluster. It contains one pod and two virtual machines to support the cluster. We'll show how easy it is to onboard on to AppDynamics, requiring minimal configuration as is needed in large-scale environments and the world of microservices, as well as show how we are able to monitor every aspect of the application as it scales to meet user demand. 00:00:37:18 - 00:01:01:10First, from the UI, we easily onboard a cloud connection to consume cloud components, including infrastructure metrics. Once that configuration is set, we can see how we pull in the infrastructure, including hosts, load balancers and storage, into the platform automatically. Reviewing the hosts, we see two hosts are being monitored through CloudWatch, including host metrics going back to investigate how many hosts are currently allocated to the Kubernetes cluster. 00:01:01:17 - 00:01:21:14We see those (same) hosts, which we see in the AppDynamics UI, are the same hosts allocated to this cluster. The cloud connection is a one-time setup. Any new hardware allocated in the cloud will be monitored automatically by AppDynamics, as we'll see later in the video. After we set up our cloud connection, we will use Helm charts to simplify the installation of monitoring components in the Kubernetes environment. 00:01:21:16 - 00:01:41:13This includes installing the app, mixed cluster operator and agent, the infrastructure and logs agents to monitor and collect logs. Additionally, we will enable the ability to auto-instrument applications using Open Telemetry, which provides a framework for capturing application telemetry data. To recap to this point, we have rapidly established a cloud connection to either a public or private cloud.Additionally, with just a few steps, we have started gathering telemetry and logs related to the Kubernetes environment. 00:01:41:16 - 00:02:04:07Next, we will auto-instrument the Pet Clinic application deployment using Open Telemetry. After playing some load to the application, the Pet Clinic service appears in AppDynamics, as shown in this flow map. We can also see the business transaction generated by this Pet Clinic service, which is coming from the Open Telemetry instrumentation. 00:02:04:10 - 00:02:21:17Here, we see three business transactions are being generated from this service. We also see the service instance of which there is only one. We see the Kubernetes cluster, the services running under, which namespace, which workloads, and how many pods of which we know there is only one deployed, and which hosts the pods are running under, of which there is only one pod. 00:02:21:23 - 00:02:39:10So, there is only one host out of the two that are running this pod. Remember, we didn't have to do any fancy configuration to enable all this.We set up a cloud connection, installed a Helm chart to enable Kubernetes monitoring, and we auto instrumented our application. We didn't have to set up an application tier or node names as we do with a commercial SaaS solution. 00:02:39:10 - 00:03:01:05And everything you see in AppDynamics was automatically ingested into the platform and automatically correlated between all entities, including APM, Kubernetes and the cloud infrastructure. Moving forward, AppDynamics will automatically monitor and show any changes to the infrastructure, such as scaling up or down, issues in Kubernetes or changes to the application, as well as any performance issues. Let’s show this by scaling the application. 00:03:01:07 - 00:03:24:05First, we see there is only one pod running for Pet Clinic service and two nodes are host to support the cluster. Now we'll scale up the Pet Clinic application service from one pod to 20 pods total. We can see all the parts started with one having an issue. So a total of 21 pods. We also see the Kubernetes cluster automatically scaled up the number of virtual hosts, EC2 instances in this case to handle the increased demand for 20 pods. 00:03:24:06 - 00:03:45:13We now have 20 pods running, one pod failed, and five hosts for the infrastructure. Coming back to the AppDynamics UI and under our Kubernetes cluster, we see AppDynamics automatically monitored and correlated the five total hosts to this cluster and Pet Clinic service. Additionally, we're also now automatically monitoring and reporting on 21 pods to support the Pet Clinic service. 00:03:45:14 - 00:04:07:09AppDynamics’ new Cloud Native solution will make it easier to onboard customer cloud environments and automatically monitor their ephemeral environments, and automatically correlate incoming MELT metrics, events, logs, and traces. Once the data is in the platform, we can apply data science, such as machine learning and AI, to help solve problems from the application, Kubernetes, infrastructure or other cloud native aspects that the customer is using.

Additional Resources

Learn more about OpenTelemetry and Kubernetes in the documentation.

Cisco AppDynamics Community

Optimizing usability for evolving applications

Intelligent monitoring in a rapidly changing digital landscape

Video Length: 4 min 17 seconds

Video Transcript

Additional Resources