Co-Authors:
IBM Observability by Instana was named a 2021 Gartner Peer Insights Customers' Choice for application performance monitoring. The solution provides enterprise observability to improve application performance management and accelerate CI/CD pipelines. Monitored applications can reside on public clouds, private clouds, hybrid clouds, on-premises environments, IBM Z®, etc.
Distributed tracing is a key capability to trace every request across services, to facilitate application troubleshooting and performance optimisation. Apache HTTPd Tracing enables you to trace your request to an Apache HTTP server, and obtain the hierarchy of upstream and downstream service calls. You can also view details of the request to an Apache HTTP Server, including HTTP header, HTTP request method, status code and the related infrastructure.
Apache HTTPd Tracing Monitoring provided by Instana offers full observability in your web applications and can immediately identify the root cause of issues that impact your service. To discover how powerful the feature is, consider the following two use cases.
1. Using Instana to monitor multiple virtual hosts on one Apache HTTPd server
Apache HTTPd Tracing provides users with the capability to monitor multiple websites running on one Apache HTTPd server in one application perspective view. Application Perspective provided by Instana is a powerful tool for monitoring, alerting, and analysis of a micro-service environment.
Suppose there are two virtual hosts to be monitored. One is shop.rumrh83-011.fyre.ibm.com, and the other is query.rumrh83-011.fyre.ibm.com. See the following virtual host definitions in the Apache HTTP server configuration file:
First of all, install instana agent and enable Apache HTTPd Tracing in your Instana agent configuration file (for more configuration details, see Apache HTTPd Tracing Configuration). You can modify the agent's configuration.xml as follows:
Then you can leverage Instana Application Perspective to monitor these two virtual hosts as two services. Click the Applications icon in the side menu, and then create a new Application Perspective to include the service shop.rumrh83-011.fyre.ibm.com and the service query.rumrh83-011.fyre.ibm.com.
Now you can get your two virtual hosts monitored in the Application Perspective view. The statistics of golden signals of these two virtual hosts are shown in the Summary dashboard.
Additionally, a dependency map is provided to display an overall architecture of your application. You can easily obtain an overview of the service dependencies, and understand the communication paths and throughput by using a visual representation of calls between services.
If you click the Analyze Calls button, you'll be redirected to the Analytics dashboard to get detailed calls or traces data about your services.
2. Using Instana to analyse erroneous calls
Suppose you see one event flagged in the Events icon in the side menu as the following screen. Click the Events icon to open the Events dashboard. Then, you may find that the event Erroneous call rate is too high listed. Click the event, and drill down to see more details.
For example, in the Event Details page, it can be seen that the erroneous call rate reached 100% in the timeline graph. Click the Analyze Calls button to go to the Analytics dashboard for more information about this erroneous calls event.
When you go to the Analytics page, you'll find that Instana already generates the appropriate Filter, Group and Chart information for the erroneous calls event. As a result of the filter, you can see the endpoint list where the metric Erroneous Calls Rate reaches 100%.
You can scroll down the endpoint list, and find the calls marked with the red flag, which means the call contains errors. Drill down to open the Call Details page.
Then, you can get an overall tree hierarchy of the calls. The erroneous call is marked with the red exclamation mark. Click the call to see more information. For example, after clicking the call, you can see that the span detail indicates a 503 Service Unavailable error is thrown, which typically means the downstream service is not available.
Summary
IBM Observability by Instana brings benefits of full observability with complex applications. You can get a real-time understanding about what happens on the Apache HTTPd server and its impact, which helps you to resolve the issues faster.
Are you ready to start the tour of Instana? Let's go...