Distributed Tracing and Bi-Directional Drilldowns

Create Tests

Create Tests

Summary #

Now we will create tests to demonstrate this integration.

We will create three tests:

• A test from the ThousandEyes agent (inside the cluster); this would be useful if your application was not acceesible from the public internet
• A more interesting trace (inside the cluster)
• A test from the public internet, using the public URL and with a ThousandEyes Cloud Agent

Step 1: HTTP Test (In Kubernetes Cluster) #

1. In ThousandEyes, go to Network & App Synthetics > Test Settings.

2. Click Add New Test and choose HTTP Server.

3. Configure the test:

   • URL: http://api-gateway.default.svc.cluster.local:82/
   • Test Name: [Name] Frontend Available (In Cluster)
   • How often test runs: 10 minutes
   • Agents: Select the Enterprise Agents tab and select the agent you deployed earlier in this guide
   • Enable distributed tracing: Enable
   • Verify Content: Optional, use PetClinic if you want to validate the returned page content.

   

4. Click Instant Test

   Instant Test

   This will open a new tab and not save your test, so be aware of this.
   

5. Switch to the Service Map tab

It may take a little time for the service map view to show up in ThousandEyes, but you should be able to find the trace in Splunk Observability Cloud

1. Copy the trace
2. In Splunk Observability Cloud, navigate to APM > Trace Analyzer, paste in the trace, and Go

Ultimately you should see something like the following:

Step 2: HTTP Test (In Kubernetes Cluster) - More interesting trace #

Now let’s repeat step 1, but using the url: http://api-gateway.default.svc.cluster.local:82/api/customer/owners (Owners List).

You can edit the same test you made, and run an instant test.

You should get more interesting maps.

ThousandEyes Failure

Notice my test in ThousandEyes failed? That’s because I didn’t change the Verify Content based on the Owners List.