Show Offer • Current London
For Apache Kafka operators

Free AI early access Kafka cluster health report

No tooling replacement required. Install the AxonOps lightweight agent alongside existing tools and in minutes you will receive a detailed report on the health of your Kafka cluster with actionable insights and remediation recommendations.

Apache Kafka Confluent Kafka
01

Install the agent

Install the lightweight AxonOps agent alongside your existing tools in minutes.

02

AI analyses

The model analyses your Kafka cluster telemetry and configuration state within minutes.

03

Get your report

Receive a focused health report with findings and remediation recommendations.

Scan to register

Claim your free Kafka health report

Register your interest now and meet the AxonOps team on the expo floor at Current London 2026.

What gets assessed

Following data points all assessed

Metrics, logs, alerts, consumer group state, configurations, topics, schemas, connectors, consumer lag, replication, and more.

Metrics and high-frequency broker telemetry

Logs, alerts, and consumer group state

Consumer lag, replication, and ISR behaviour

Topics, partitions, schemas, and connector estate

Configuration drift across brokers and clients

Existing tooling stays in place while AxonOps works alongside it

Example Health Report Summary

A report that points to the issue, not just the symptom

The goal is not another generic dashboard. The report highlights what is unhealthy, why it matters operationally, and what to do next.

Cluster health report • prod-eu-west-1
Generated 00:04:37 after install
Brokers 12 5s scrape
Topics 347
Partitions 6,184
Findings 3 / 7 / 8
Critical

Consumer lag growing on payments-events, partition 7

Group fraud-scoring-v3 is 1.4M messages behind on broker-3 and has grown linearly for 22 minutes. Root cause points to a single slow consumer holding back rebalance progress.

Restart consumer fraud-scoring-v3-pod-7 so rebalance can redistribute partition 7 within 30 seconds.
Warn

Under-replicated partitions on user-clicks

28 partitions have ISR 2 instead of target 3 since broker-9 rejoined the cluster. Replica fetcher throttle is currently capped at 1 MB/s which is too low for catch-up before the next retention sweep.

Raise follower.replication.throttled.rate to 50 MB/s on broker-9 until under-replicated partitions return to zero.
Info

ISR shrink / expand storms on inventory-updates

42 ISR transitions in the last hour all on partitions led by broker-5. JVM GC pauses on that broker correlate with the pattern and indicate an instability precursor.

Tune G1GC MaxGCPauseMillis=200 on broker-5 and schedule a rolling restart this week.