Browse Kafka Consumer Groups and Assignments
Kafma lists the consumer groups known to the connected cluster. Use the list to find a group, check whether it is active, and inspect its assignments and members from the same page.
What is a Kafka consumer group?
A Kafka consumer group is a set of consumers that share a group ID and divide topic partitions between them. Within the group, each assigned partition is read by one member at a time. Adding members can increase parallelism up to the number of partitions, provided those members subscribe to the topic.
Different consumer groups read the same topic independently. Each group keeps its own committed position for every partition, so one application can be caught up while another is still processing older records.
What is a Kafka consumer offset?
An offset identifies a position in one Kafka partition. Because every partition has its own offset sequence, an offset is meaningful only with its topic and partition.
A consumer group commits offsets to record where it should continue. A committed offset normally points to the next record the group should read, not the last record it processed. Kafma shows these committed offsets beside each partition's end offset, making it possible to inspect assignments and calculate consumer lag.
List and filter Kafka consumer groups
Search by group ID or filter by Stable, Rebalancing, Empty, or Dead. Each row shows:
- The group protocol.
- Its state and active member count.
- Its topic and partition assignment.
- Total lag and the number of lagging partitions.
Refresh the list when membership or assignments change. Pin frequently checked groups to the Watchlist for quicker access from Home.
Inspect members and topic assignments
Select a group to open its details in a drawer without leaving the Consumer Groups page. The summary shows its state, member count and assignor, total lag, maximum partition lag, and recent lag trend.
Expand a topic to see each partition's committed offset, lag, and current assignment. The Members section maps member IDs and hosts to their topic and partition assignments, which helps locate the process handling a partition.
Use the Console action beside a topic, or hover over an assigned partition and select Inspect, to open Watch Group at the group's committed positions. This connects assignment data to the messages the consumer is approaching.
Understand group state and protocol
| Value | Meaning |
|---|---|
| Stable | The group has completed assignment and is not rebalancing |
| Rebalancing | Members or assignments are currently changing |
| Empty | The group has no active members but may still have committed offsets |
| Dead | The coordinator no longer considers the group active |
| Unknown | Kafka did not return a recognized state |
The badge is Kafma's classification of the protocol type returned by Kafka.
CLASSIC means Kafka reported the standard consumer protocol. SIMPLE
covers other protocol types. These groups can have saved offsets without a
standard consumer assignment, so Unassigned is not necessarily an error.
State and lag answer different questions. A Stable group can still be behind, while an Empty group can show zero lag if no new records have arrived.
Delete a Kafka consumer group
Stop all consumers before deleting a group. Kafka rejects deletion while the group is active.
Deleting a group removes the group and its committed offsets. It does not delete
topics or messages. If the application later starts with the same group ID,
Kafka has no committed offset to return. The consumer then applies
auto.offset.reset for each partition: earliest starts at the oldest retained
offset, latest starts at the current end, and none fails instead of choosing
a position.
Write actions are unavailable on read-only connections.