Kafka graphs

Use Kafka graphs to monitor producer and consumer performance in Kafka-based systems.

Overview

Kafka graphs provide dedicated monitoring of Kafka message throughput, producer duration, and message rates within the dashboard. These metrics help you understand how Kafka operations contribute to overall application performance.

You can correlate Kafka metrics with transaction performance, system resources, and other application data to identify bottlenecks and optimize performance.

Kafka metrics

The following table describes the available Kafka metrics.

Metric Description Applies to

Kafka Producer Duration

Time taken to send messages to Kafka. Measured in milliseconds. High producer duration may indicate network latency or broker congestion.

Producer

Kafka Throughput

Number of messages processed per second. Measured in messages/sec. Helps identify message processing capacity and constraints.

Producer and Consumer

Kafka Message Rate

Number of messages processed over time. Displays message activity trends and patterns during the run.

Producer and Consumer

Access Kafka graphs

View Kafka metrics on the dashboard.

To access Kafka graphs:

  1. Open the dashboard for a run.

  2. In the Metrics tree, locate the Kafka category.

  3. Select the desired metrics. The graph displays data for the run.

    For details on configuring graph display options, see Graph display options.

Correlate Kafka with application performance

Kafka metrics are most valuable when viewed alongside other performance data. Use the following guidance to interpret Kafka metrics in context:

  • High producer duration + high transaction response time: Kafka operations may be contributing to application slowness. Investigate broker health and network latency.

  • High throughput + stable producer duration: Kafka is processing messages efficiently; look for bottlenecks in other parts of the application.

  • Low message rate + degraded system resources: System resource constraints may be limiting Kafka message processing capacity.

  • Inconsistent message rate + error spikes: Errors may be causing message processing delays. Check the Errors tab for related errors.

Use the dashboard to place Kafka graphs on the same tab as transaction graphs, system resource graphs, and error summaries to enable side-by-side correlation analysis.

See also