Skip to main content
The Insights dashboard provides comprehensive analysis of your Kafka infrastructure, helping platform teams to:
  • proactively identify configuration issues,
  • optimize storage costs,
  • track governance metrics,
  • monitor business-critical topics and more.
To access the dashboard, click Insights on the left menu in Conduktor UI. Insights dashboard

Overview

Insights analyzes your Kafka cluster data and presents actionable intelligence across four key areas:

Metrics summary

At the top of the dashboard you’ll see a cluster-wide summary of all the metrics:
  • Topics - the number of topics in the cluster
  • Partitions - an aggregate partition count across all topics
  • Consumed - the total number of consumer groups consuming from topics
  • Topic distribution - a breakdown of topics by type (internal, streams or user)
  • Health score - the cluster health indicator based on our risk analysis
These metrics update dynamically, based on applied filters.

Filter data

Use the filter at the top of the Insights dashboard to narrow down data across all sections. Filters apply globally to the summary metrics, risk analysis, cost control, VIP topics and governance sections. Filter topics by their classification:
  • Internal - internal topics (e.g., __consumer_offsets)
  • Streams - Kafka Streams internal topics (e.g., changelog, repartition topics)
  • User - application topics created by you
You can also filter topics by labels. Select one or more labels to only see topics with matching labels across all the Insights sections.
Clicking a label in any table also applies it as a filter, making it easy to explore related topics.
Click Clear all to remove all applied filters and return to the full cluster view.

Export data

You can export data as a .zip file containing CSV files. Choose whether to export all available Insights data or only section-specific metrics. When filters are applied, the export includes only the filtered data and file names include a -filtered suffix (e.g., risk-analysis-filtered.csv) to indicate the data is a subset of the full cluster. The exported data can be shared with stakeholders, used in offline analysis, tracked over time to support capacity planning decisions or archived for audit purposes.