Why clustering provides horizontal scaling in Redis - Visual Breakdown

Concept Flow - Why clustering provides horizontal scaling

Start with single Redis node

↓

Add more Redis nodes (cluster)

↓

Data split across nodes

↓

Requests distributed to nodes

↓

Load shared, faster response

↓

System scales by adding nodes

↓

Horizontal Scaling

Clustering splits data and requests across multiple Redis nodes, sharing load and allowing the system to grow by adding nodes.

Execution Sample

Redis

1. Start with 1 Redis node
2. Add node 2 and node 3
3. Data partitions assigned to nodes
4. Client requests routed to correct node
5. Load balanced, faster processing

Shows how adding nodes and splitting data leads to horizontal scaling in Redis clustering.

Execution Table

Step	Action	Data Distribution	Request Handling	Effect on Load
1	Single Redis node running	All data on node 1	All requests to node 1	Load on single node
2	Add node 2 and node 3	Data still on node 1	Requests still to node 1	Load unchanged
3	Data partitioned across nodes	Data split: node1, node2, node3	Requests routed to correct node	Load shared across nodes
4	Client sends request for key in node2	Key located on node2	Request handled by node2	Load on node2 increases, others less busy
5	Add node 4	Data repartitioned to include node4	Requests routed accordingly	Load further balanced, system scales
6	System handles more requests	Data and requests spread	Parallel processing on nodes	Faster response, horizontal scaling
7	Exit	N/A	N/A	Scaling achieved by adding nodes

💡 System stops adding nodes when desired scale or resource limit reached

Variable Tracker

Variable	Start	After Step 2	After Step 3	After Step 5	Final
Number of Nodes	1	3	3	4	4
Data Distribution	All on node1	All on node1	Split across 3 nodes	Split across 4 nodes	Split across 4 nodes
Request Routing	To node1	To node1	To correct node	To correct node	To correct node
Load per Node	High on node1	High on node1	Shared across 3 nodes	Shared across 4 nodes	Shared across 4 nodes

Key Moments - 3 Insights

Why doesn't adding nodes immediately reduce load before data is partitioned?

Because initially data and requests remain on the original node (see execution_table step 2), so load is unchanged until data is split.

How does Redis know which node handles a request?

Redis uses data partitioning (hash slots) to route requests to the node holding the key (see execution_table step 4).

Why is adding nodes called horizontal scaling?

Because it adds more machines (nodes) side-by-side to share load, unlike vertical scaling which upgrades a single machine (see concept_flow).

Visual Quiz - 3 Questions

Look at the execution_table at step 3, how is data distributed?

A Data is split across node1, node2, and node3

B Data is duplicated on all nodes

C All data remains on node 1

D Data is removed from node1

💡 Hint

Check the 'Data Distribution' column at step 3 in execution_table

At which step does request routing start directing requests to different nodes?

A Step 2

B Step 3

C Step 1

D Step 5

💡 Hint

Look at 'Request Handling' column in execution_table to see when routing changes

If we add more nodes but do not repartition data, what happens to load?

A Load is shared evenly

B Load increases on all nodes

C Load remains high on original node

D Load disappears

💡 Hint

Refer to step 2 in execution_table and variable_tracker for load per node

Concept Snapshot

Redis clustering splits data across multiple nodes.
Requests go to the node holding the data.
Adding nodes shares load horizontally.
This improves performance and capacity.
Horizontal scaling means adding machines side-by-side.

Full Transcript

This visual execution shows how Redis clustering enables horizontal scaling. Initially, a single Redis node holds all data and handles all requests, causing high load on one machine. When more nodes are added, data is still on the original node, so load does not reduce immediately. Once data is partitioned across nodes, requests are routed to the correct node holding the data. This spreads the load across multiple machines. Adding more nodes and repartitioning data further balances load and allows the system to handle more requests efficiently. This process is called horizontal scaling because it adds more machines side-by-side to share the workload, unlike vertical scaling which upgrades a single machine. The execution table traces each step, showing data distribution, request routing, and load effects. The variable tracker follows key variables like number of nodes and load per node. Key moments clarify common confusions about when load reduces and how routing works. The quiz tests understanding of data distribution and request routing steps. Overall, clustering in Redis provides a way to grow capacity by adding nodes and splitting data and requests among them.