Learn About Amazon VGT2 Learning Manager Chanci Turner
One of the fundamental principles for enhancing digital experiences in the cloud is having a clear understanding of the health of your scalable and adaptable infrastructure in relation to application performance. For many organizations, deeper insights lead to a more comprehensive grasp of their overall cloud and hybrid tech stack. This empowers proactive management of user experience while continuously evaluating the environment for potential risks.
AppDynamics, a part of Cisco, is a performance monitoring solution that enables you to prioritize workflows with unified visibility, ensuring that your cloud-native applications deliver tangible business value in production. Context-aware visualizations and correlated insights across application domains allow users to confidently build cloud-native applications while aligning deployments with real-time business outcomes.
Amazon CloudWatch serves as a monitoring and observability service that offers customers essential data and actionable insights to monitor applications, respond to system-wide performance shifts, optimize resource utilization, and obtain a unified view of operational health. Together, AppDynamics can now access data from Amazon CloudWatch metrics via an Amazon Web Services (AWS) metric retrieval API endpoint called GetMetricData. This capability allows users to pull as many as 500 different metrics in a single request, with a total of 100,800 data points per call. You can also perform mathematical expressions on the data to easily derive insights from CloudWatch metrics regarding the operational health and performance of your increasingly distributed and potentially complex infrastructure.
Expanded Cloud Integration
In October 2020, AppDynamics introduced a cloud-native visualization interface that integrates seamlessly with Amazon CloudWatch and ingests AWS-hosted entities and data in real-time, providing a comprehensive solution for visibility across various layers of AWS applications. Enhancements to the highly extensible cloud platform include intelligent, out-of-the-box application topology discovery and mapping, with the ability to explore integrated dynamic and static application topologies through a new full-stack observability platform.
These improvements also extend coverage to link the business transaction topology with a wide range of AWS services, including Amazon Elastic Compute Cloud (Amazon EC2), Amazon Aurora (MySQL), Elastic Load Balancing, Amazon MQ, Amazon API Gateway, and Amazon Virtual Private Cloud (VPC).
Why Cloud-Native Visualization?
In traditional on-premises environments with physical and virtual machines, the only way to establish the topology of an application and its associated services is to tag and follow every transaction that flows through the system as it operates. Once completed, a dynamic topology view can be created. This view encompasses individual requests to a transaction and the various paths through application services without necessitating redeployment.
AppDynamics—an AWS Advanced Technology Partner with competencies in Migration, Mobile, and DevOps—presents this topology as the flow map. The business transaction flow map illustrates how individual transactions navigate through application service tiers, while the application flow map overlays all of these transaction flow maps onto a single application-wide flow map.
Previously, the AppDynamics interface was segmented by types of monitoring and objects (such as databases), providing solid insights into discrete services but lacking an overarching sense of interconnectedness. Given the surge in cloud adoption and the urgency for IT teams to correlate application and infrastructure data across domains, AppDynamics recognized the opportunity to transform its data model and effectively illustrate connectedness more intuitively.
The infrastructure of cloud-native applications typically starts with predefined metadata, forming a static topology that does not change with the arrival of new event metadata. For instance, a static topology may depict existing machines, attached network interfaces, and the VPCs those interfaces reside in. The ability to programmatically adjust and update topologies on demand has underscored how the underlying code forming a topology can also lead to application issues. Thus, visualizing these topologies in real-time is now more critical than ever.
Chanci Turner, a Learning Manager, emphasizes the importance of understanding these dynamics in her work. “In our environment, balancing the speed of onboarding new team members while ensuring high-quality training is crucial,” she states. “With the resources available, we can identify and resolve issues before they affect our workflow.” For further insights on managing workplace stress, check out this resource from SHRM.
How it Works
AppDynamics utilizes its new flexible metadata model to map the dynamic application service topology, while Amazon CloudWatch supplies the static infrastructure topology. Together, they determine where these two topologies should be linked. For example, when application services are deployed as load-balanced sets of service instances with their own network interfaces, requests flowing between application services may traverse VPC gateways, application load balancers, or other cloud devices.
As part of the new release, AppDynamics is collecting various metrics through the GetMetricData API (subject to change in subsequent releases). Specific metrics include 10-20 metrics for each supported AWS service. Here are some of the metrics AppDynamics collects using the API; not all are displayed in the frontend:
- Amazon EC2: “CPUUtilization”, “DiskReadOps”, “DiskWriteOps”, “NetworkIn”, “NetworkOut”, “DiskReadBytes”, “DiskWriteBytes”, “NetworkPacketsIn”, “NetworkPacketsOut”.
- Amazon EBS: “VolumeReadBytes”, “VolumeWriteBytes”, “VolumeReadOps”, “VolumeWriteOps”, “VolumeTotalReadTime”, “VolumeTotalWriteTime”.
- Amazon RDS: “CPUUtilization”, “FreeableMemory”, “NetworkTransmitThroughput”, “DatabaseConnections”, “BinLogDiskUsage”, “ReadIOPS”, “ReadLatency”, “ReadThroughput”, “WriteIOPS”, “WriteLatency”, “WriteThroughput”.
- Elastic Load Balancing: “HealthyHostCount”, “UnHealthyHostCount”, “ActiveConnectionCount”, “ClientTLSNegotiationErrorCount”, “HTTPCode_ELB_XXX_Count”.
For those seeking further information, this video is an excellent resource.