Consumer groups concept in Redis - Step-by-Step Execution

Concept Flow - Consumer groups concept

Messages added to stream

↓

Create consumer group

↓

Consumers read messages

↓

Messages assigned to consumers

↓

Consumers acknowledge messages

↓

Pending messages tracked

↓

Reassign unacknowledged messages if needed

Messages enter a stream, a consumer group is created, consumers read and acknowledge messages, and pending messages are tracked for reliability.

Execution Sample

Redis

XGROUP CREATE mystream mygroup 0
XREADGROUP GROUP mygroup Alice COUNT 2 STREAMS mystream >
XACK mystream mygroup 1526985058136-0

Create a consumer group, read messages as a consumer, then acknowledge processed messages.

Execution Table

Step	Action	Input/Command	Result/State Change	Output
1	Add messages to stream	XADD mystream * name Alice	Message added with ID 1526985058136-0	1526985058136-0
2	Create consumer group	XGROUP CREATE mystream mygroup 0	Group 'mygroup' created starting from the first message	OK
3	Consumer reads messages	XREADGROUP GROUP mygroup Alice COUNT 2 STREAMS mystream >	Consumer 'Alice' assigned new messages	Messages with IDs 1526985058136-0
4	Consumer acknowledges message	XACK mystream mygroup 1526985058136-0	Message marked as processed, removed from pending	1
5	Check pending messages	XPENDING mystream mygroup	No pending messages for 'Alice'	Empty list
6	Exit	No more messages	No new messages to assign	End of processing

💡 No new messages available, consumer group processing ends

Variable Tracker

Variable	Start	After Step 1	After Step 2	After Step 3	After Step 4	Final
Stream 'mystream'	empty	1 message (1526985058136-0)	1 message	1 message assigned to Alice	1 message acknowledged (removed from pending)	1 message in stream, none pending
Consumer group 'mygroup'	none	none	created	active with Alice assigned message	message acknowledged	active, no pending messages
Pending messages	none	none	none	1 message pending for Alice	0 pending	0 pending

Key Moments - 3 Insights

Why does the consumer group start reading from '0' when created?

Starting from '0' means the group reads messages from the beginning of the stream, as shown in Step 2 and Step 3 of the execution_table.

What happens if a consumer does not acknowledge a message?

The message stays in the pending list (Step 4 and 5), so it can be reassigned or retried later, ensuring no message is lost.

Why do we use '>' in the XREADGROUP command?

'>' tells Redis to deliver only new messages that were never delivered to any consumer, as seen in Step 3.

Visual Quiz - 3 Questions

Look at the execution_table, what is the state of pending messages after Step 3?

A One message pending for consumer Alice

B No pending messages

C All messages acknowledged

D Pending messages for multiple consumers

💡 Hint

Check the 'Result/State Change' column in Step 3 and Step 4

At which step does the consumer group get created?

A Step 1

B Step 3

C Step 2

D Step 4

💡 Hint

Look for the 'Create consumer group' action in the execution_table

If the consumer never acknowledges messages, what would happen to the pending messages?

A They get deleted immediately

B They remain pending and can be reassigned

C They are delivered to other consumers instantly

D They disappear from the stream

💡 Hint

Refer to the key_moments explanation about unacknowledged messages

Concept Snapshot

Consumer groups let multiple consumers share a stream.
Create a group with XGROUP CREATE.
Consumers read with XREADGROUP using '>' for new messages.
Consumers acknowledge messages with XACK.
Pending messages track unacknowledged messages.
This ensures reliable, distributed message processing.

Full Transcript

In Redis streams, messages are added to a stream. A consumer group is created to allow multiple consumers to read from the stream without duplicating work. When a consumer reads messages using XREADGROUP with the '>' symbol, it receives new messages not yet delivered to any consumer. After processing, the consumer acknowledges messages with XACK, removing them from the pending list. Pending messages are tracked to handle cases where consumers fail to acknowledge, allowing reassignment and ensuring no message is lost. This flow supports reliable and scalable message consumption.