Instana U

 View Only

Observe a sample microservices application running on Kubernetes using Instana

By IRFAD K P posted Wed February 28, 2024 08:50 AM


IBM Instana is a fully automated real-time observability platform that contextualizes performance data, enabling quick identification to prevent and address issues promptly. IBM Instana Observability goes beyond typical Application Performance Monitoring (APM) solutions, making observability accessible to everyone in DevOps, SRE, platform engineering, ITOps, and development. It also ensures that users can access the data they need with the necessary context. 

For monitoring purposes, an expanded version of a sample microservice application known as Robot Shop used to observe a typical microservice application via Instana. This application incorporates a diverse range of technologies, including NodeJS (Express), Java (Spring Boot), Python (Flask), Golang, PHP (Apache), MongoDB, Redis, MySQL (Maxmind data), RabbitMQ, Nginx, and AngularJS (1.x). The application itself is deployed within a Google Kubernetes Engine (GKE) cluster. 

The initial step involves the installation of the Instana agent onto the Kubernetes cluster where Robotshop is deployed. Upon installation, Instana automatically detects the various technologies that are employed within the environment, and provides accessible data for monitoring purposes. 

For more information on installing the Instana agent across different platforms, see Installation page by going to Agent > install an Agent. 



After the installation, you can monitor the data in your Dashboard as seen in the following screenshot: 



The following sections cover some of the core Instana features: 

Attain comprehensive visibility across your IT landscape 

Gaining visibility into containers can be challenging. However, with Instana, you have the flexibility to seamlessly switch between hosts and containers, empowering you to choose the perspective that best suits your needs. 

To attain comprehensive visibility, go to the left panel, click Infrastructure and then select the Perspective and Grouping filter to see your infrastructure in different aspects.  



Instana comprehensively grasps the interactions among the components within your infrastructure. Through a dashboard view, you can visually comprehend the placement and integration of each container within the stack, providing a clear understanding of their interrelationships. 

 To see the data of a specific host as seen in the following screenshot, click on any of the hosts and then click Open Dashboard:  


Now, you can see that the prebuilt dashboards provide you with comprehensive and relevant information about your environment, offering a consolidated view of crucial data points and insights. 




Instana infrastructure monitoring employs AI-powered automation for monitoring, alerting, and remediation. It offers unparalleled, real-time visibility into intricate, distributed applications, services, and various infrastructure components such as servers, containers, and databases. This comprehensive approach ensures that teams haveall necessary data and context to proactively prevent downtime, optimize resource usage, and improve overall user experience and productivity.  For more information, see Infrastructure monitoring. 

Figure out why you have a load time latency in the website

Click the Websites tab, and then click on the Robot Shop website as seen in the following screenshot: 



Here, you can view data about the robotshop website and its latency details. To assess the potential impact on the overall onLoad time, examine the resource load times. Similarly, to identify which image is causing latency in load time, review the details in the Load Time tab. 



Instana automatically gathers and contextualizes crucial details for each web browser request, ensuring thorough performance insights. This monitoring solution utilizes a lightweight JavaScript agent seamlessly integrated into the monitored website to facilitate these functions. Smarts alerts within the website with pre-configured alert setups based on out-of-the-box blueprints. These blueprints cover scenarios like website slowness, JavaScript errors, and HTTP status codes. By selecting a blueprint for alerting and specifying criteria such as geolocation, web browser, or operating system, the system generates customized alerts. This ensures a continuous monitoring process to uphold an optimal website experience. For more information, see Website monitoring. 


Check performance for services your own 

The Application Perspective is a powerful too that actively monitors and analyses analyzing microservice environments. The tool is designed to provide real-time alerts, serving as a proactive mechanism to efficiently manage and address any potential issues that may arise. 


The functionality of the Robot Shop application relies on the seamless collaboration of numerous interdependent services. 



 The Dependencies graph shown in the following screenshot provides a visual representation of the interactions among services within the eCommerce application, illustrating how users engage with the platform during their shopping experience. For a comprehensive analysis of the traffic flow, click any of the services displayed.



 This action reveals both upstream and downstream dependencies that are associated with the selected service. 


 An Application Perspective (AP) serves as a robust utility for overseeing, signaling, and scrutinizing a microservices ecosystem. Each AP autonomously creates an elaborate monitoring interface for critical performance indicators and additional metrics. It facilitates team alignment, enabling them to concentrate on their pertinent services without diversions. The containment of alerts, errors, and logs within an AP framework streamlines the diagnostic process. 


Also, from a security standpoint, organizations can use APs to restrict access to infrastructure and services, enhancing overall visibility control. For more information, see Application perspective. 

Solve an incident 

 The incidents are identifiable by yellow or red icons, while the disabled ones are marked in gray. Anomalous calls that affect the Robot Shop application have been observed. To begin the investigation process, select the relevant incident. 



 A triggering event might not be the initial event in a sequence, but it is the one responsible for causing a decline in service or application performance: 



 Failed calls are associated with multiple groups. It might be necessary to investigate across all these groups to address the issue effectively. 


 Upon delving into a specific group, you gain visibility into all the problematic calls located within that particular group. This streamlined approach offers a more convenient way to gain comprehensive insight into the issues at hand. 



 An incident helps to understand the conditions that affect both your edge services and vital infrastructure by learning their patterns and conditions. It notifies you when issues occur by recognizing abnormalities and sending alerts. Edge services are the accessible parts of your application that customers or external systems use, and they act as the external interface of the application. For more information, see Event types