Troubleshooting Traces

If you have issues with trace data collection, the following sections can help you identify and troubleshoot the issue.

I don’t see anything on the Traces page

The Traces page may not show anything if LogicMonitor is not receiving or properly processing the traces forwarded from your applications. In this case, you can perform the following checks to isolate the issue and identify the cause:

Are there Errors in LogicMonitor audit logs?

Search the audit log for the user you used to set your OpenTelemetry Collector. If the Collector is sending traces to LogicMonitor, you will see an audit log entry for the creation of the service.

Does the OpenTelemetry Collector receive spans?

LogicMonitor monitors each OpenTelemetry Collector that is added to automatically detect traces. You can use these metrics to verify if traces are being received from the application to the OpenTelemetry Collector and forwarded to your LogicMonitor account. If you don’t see the Receiver and Exporter DataSources, the Collector hasn’t received any spans form your application.

Duplicate Resources are being created

For each trace that is forwarded, LogicMonitor tries to map the ‘host’ attributes to an existing monitored resource. (The ‘host’ attribute is not required by OpenTelemetry.)

• If LogicMonitor does not find a matching ‘host’ resource, it creates a new resource.
• If the ‘host’ attributes are not specified, a service will be automatically created based on the ‘service’ attribute and all trace data will mapped to the service.

If you instrumented your applications to send ‘host’ attributes with the traces forwarded to LogicMonitor, but you’re not seeing the trace data associated to the correct monitor resources, make sure that following values are provided and current during instrumentation:

host.name=(hostname mapped to system.hostname in LM)
ip=(ip address of the resource mapped to system.hostname in LM)
resource.type={pod, cloud, physical}
resource.group={group name or id for the resource}

Note: For Kubernetes, we recommend that you add your cluster into monitoring before forwarding traces from your cluster to ensure that traces are associated with the correct pods and that duplicate resources are not created.

I want to change how Traces are displaying

The Traces page does not let you modify how traces are displayed. But if you have a problem with how traces are displaying, you will need to make the change on the instrumentation.

• If you want to change the name of a service, change the parameters for your automatically instrumented application.
• If you want to change the name of an operation, change the name from your manual instrumentation.
• If your operations appear to be out of order in the end-to-end trace view, make sure the NTP is synced for your host.