KAFKA MESSAGE BROWSER

Find the Kafka record.
Inspect its full context.

Browse Kafka records by partition, offset, or time; combine bounded filters, follow a live tail, inspect complete context, and safely publish a new immutable record.

See the workflow Read the API guide

Product UI · sample data

TOOLS

Message explorer

Search, inspect, compare, export, and safely republish Kafka records.

Recordsorders.events.v1 · bounded snapshot

Live ready

Matched248of 1,000 scanned

Partitions12 / 12full sample

Window14m 58sevent time

Rate38.4/soffset delta

Messages over timeLast 15 min

Partition activityRecords

Keyorder-10482Offset18,442,901

{
  "eventId": "evt_demo_01J8K4R2",
  "orderId": "ord_demo_84271",
  "customerId": "cus_demo_1092",
  "status": "created",
  "currency": "EUR",
  "amount": 129.50
}

How search works: BetterKafka scans within explicit partition, record-count, time and concurrency bounds. It does not claim a broker-side full-text index or unlimited message history.

INVESTIGATION WORKFLOW

From broad signal to exact context.

Search without guesswork

Combine JSON-path, header, metadata, key, value and timestamp conditions. Safe timed regular expressions, scan telemetry and partition watermarks keep the scope visible.

Forward and backward cursors
Saved browser searches
JSON and NDJSON export

Inspect in context

Review keys, values, duplicate headers, offsets, timestamps, encodings, masks and truncation indicators without hiding the Kafka record model.

UTF-8 and base64-safe display
Partition/time distribution charts
Explicit scan coverage

Publish deliberately

Kafka records remain immutable. BetterKafka composes a new record, republishes with changes, or sends a keyed tombstone when the selected cluster and user role permit it.

Single and NDJSON batches
Validation preview and receipts
Operator reasons for batches

LIVE TAIL

Watch a bounded stream without committing offsets.

The SSE tail supports pause and stop semantics, record and duration limits, duplicate suppression, reconnect handling, and per-window charts. BetterKafka reads for inspection; it does not join your consumer group.

Live tail3,240 records/s

Partition 7Window: 15 minutesLimit: 1,000 records

FIND THE MESSAGE

Make record investigations repeatable.

Deploy BetterKafka with your own response-masking rules and cluster permissions.

Start free trial Install with Helm