⚙️ What is Redis Streams?

Definition:

Redis Streams is a log-based data structure in Redis, designed for storing, processing, and consuming real-time streams of data.

• Think of it as a persistent, append-only message queue.
• Supports multiple producers and multiple consumers.
• Useful for event sourcing, message queues, data pipelines.

Analogy:

• Like a Kafka topic, but fully managed in Redis.

---

🔄 Key Concepts

Term	Description
Stream	The data structure itself (like a topic)
Entry	Individual message/event in the stream
ID	Unique identifier for each entry (timestamp-based)
XADD	Command to add an entry to a stream
XRANGE / XREVRANGE	Read entries from a stream
Consumer Group	Group of consumers that share messages
XREAD / XREADGROUP	Commands to consume messages
Pending Entries List (PEL)	Tracks messages delivered but not acknowledged

---

🔧 Basic Commands

1. XADD – Add a message to the stream

XADD mystream * user "Alice" action "login"

• * → auto-generated ID based on timestamp
• user and action → fields of the message

1. XRANGE – Read messages by range

XRANGE mystream - +

• - → start (oldest)
• + → end (newest)

1. XREAD – Read messages from one or more streams

XREAD COUNT 10 STREAMS mystream 0

1. Consumer Groups – multiple consumers share the load

XGROUP CREATE mystream mygroup $ MKSTREAM
XREADGROUP GROUP mygroup Alice COUNT 10 STREAMS mystream >

• > → fetch new messages
• Each consumer in the group gets different messages (like queue semantics)

1. Acknowledgement (XACK) – mark message as processed

XACK mystream mygroup 1696890000000-0

---

🔄 Features of Redis Streams

1. Persistent Storage

   • Unlike normal Pub/Sub, streams store messages, so consumers can read later.

2. Multiple Consumers & Consumer Groups

   • Scale processing horizontally.

3. Automatic IDs

   • Unique, time-ordered IDs for each entry.

4. Message Acknowledgment & PEL

   • Track pending messages → ensures exactly-once processing in consumer groups.

5. Efficient Range Queries

   • Read entries by ID range → supports time-based processing.

---

🧩 Example Use Case

Scenario: Logging user actions in real-time:

1. Microservices produce events:

   • user_signup, user_login, order_placed
2. Each event is XADDed to a Redis stream: XADD user-events * type "login" user_id "1234"

3. Consumer group processes events:

   • Service A → Analytics
   • Service B → Notifications
4. Messages are acknowledged after processing (XACK)

• Guarantees no message is lost, multiple consumers can process independently.

---

⚡ Redis Streams vs Pub/Sub

Feature	Redis Pub/Sub	Redis Streams
Persistence	❌ Messages lost if no subscriber	✅ Messages stored in stream
Consumer Groups	❌ Not supported	✅ Supported
Replay	❌ No	✅ Yes, using IDs
Exactly-once	❌ Hard	✅ Using PEL & XACK
Use Case	Real-time notifications	Event sourcing, queues, pipelines

---

🧭 Key Takeaways

• Redis Streams = log-based, persistent data structure for real-time events.
• Supports producer-consumer model with consumer groups.
• Guarantees message durability, replay, and acknowledgment.
• Great for real-time pipelines, microservices communication, and analytics events.

---

💡 Analogy:

• Streams = train track, each entry = passenger.
• Consumers = stations, PEL = tickets of passengers not yet delivered.

---