Rebalancing behavior in Kafka - Step-by-Step Execution

Process Flow - Rebalancing behavior

Consumer Group Starts

↓

Coordinator Assigns Partitions

↓

Consumers Receive Assignments

↓

Consumers Process Messages

↓

Consumer Joins/Leaves or Session Timeout

↓

Trigger Rebalance

↓

Coordinator Reassigns Partitions

↓

Consumers Receive New Assignments

↓

Continue Processing

This flow shows how Kafka consumer groups rebalance partitions when consumers join, leave, or fail.

Execution Sample

Kafka

consumer1 joins group
coordinator assigns partitions
consumer1 processes assigned partitions
consumer2 joins group
rebalance triggered
coordinator reassigns partitions
consumers receive new assignments

Simulates consumers joining a group and Kafka rebalancing partition assignments.

Process Table

Step	Event	Coordinator Action	Consumer State	Partition Assignment
1	consumer1 joins group	Assign all partitions to consumer1	consumer1 connected	consumer1: [partition0, partition1, partition2]
2	consumer1 processes messages	No change	consumer1 processing	consumer1: [partition0, partition1, partition2]
3	consumer2 joins group	Trigger rebalance	consumer1 and consumer2 connected	none (rebalancing)
4	rebalance triggered	Reassign partitions evenly	consumer1 and consumer2 paused	none (rebalancing)
5	coordinator reassigns partitions	Assign partitions to consumers	consumer1 and consumer2 ready	consumer1: [partition0, partition1], consumer2: [partition2]
6	consumers resume processing	No change	consumer1 and consumer2 processing	consumer1: [partition0, partition1], consumer2: [partition2]
7	consumer1 leaves group	Trigger rebalance	consumer2 connected	none (rebalancing)
8	rebalance triggered	Assign all partitions to consumer2	consumer2 ready	consumer2: [partition0, partition1, partition2]
9	consumer2 processes messages	No change	consumer2 processing	consumer2: [partition0, partition1, partition2]

💡 Rebalancing stops when all consumers have stable partition assignments.

Status Tracker

Variable	Start	After Step 3	After Step 5	After Step 7	Final
consumer1_state	not connected	connected	processing	left group	left group
consumer2_state	not connected	connected	processing	processing	processing
partition_assignment	none	none (rebalancing)	consumer1: [partition0, partition1], consumer2: [partition2]	none (rebalancing)	consumer2: [partition0, partition1, partition2]

Key Moments - 3 Insights

Why do consumers stop processing during rebalance?

During rebalance (see steps 3-5 in execution_table), consumers pause processing to avoid duplicate processing and ensure partitions are reassigned cleanly.

What triggers a rebalance in a consumer group?

A rebalance is triggered when a consumer joins or leaves the group or when a consumer session times out, as shown at steps 3 and 7.

How are partitions assigned after rebalance?

The coordinator evenly distributes partitions among active consumers, as seen in step 5 where partitions are split between consumer1 and consumer2.

Visual Quiz - 3 Questions

Look at the execution_table, what is the partition assignment after step 5?

A consumer2: [partition0, partition1, partition2]

B consumer1: [partition0, partition1, partition2]

C consumer1: [partition0, partition1], consumer2: [partition2]

D No partitions assigned

💡 Hint

Check the 'Partition Assignment' column at step 5 in execution_table.

At which step does consumer2 start processing messages?

A Step 6

B Step 3

C Step 7

D Step 9

💡 Hint

Look for when consumer2 state changes to processing in execution_table.

If consumer1 never leaves the group, what would happen to partition assignments after step 7?

A All partitions assigned to consumer2

B Partitions remain assigned as in step 5

C Rebalance triggered again

D No partitions assigned

💡 Hint

Refer to variable_tracker for consumer1_state and partition_assignment after step 7.

Concept Snapshot

Kafka consumer groups assign partitions to consumers.
When consumers join or leave, a rebalance happens.
During rebalance, consumers pause processing.
Coordinator reassigns partitions evenly.
Consumers resume processing after rebalance.

Full Transcript

This visual execution shows how Kafka consumer groups handle rebalancing. When a consumer joins or leaves, the coordinator triggers a rebalance. During rebalance, consumers pause processing to avoid duplicates. The coordinator then reassigns partitions evenly among active consumers. After assignments, consumers resume processing. The execution table traces each step, showing consumer states and partition assignments. Key moments clarify why processing pauses and what triggers rebalances. The quiz tests understanding of partition assignments and rebalance triggers.