Worker processes and connections in Nginx - Step-by-Step Execution

Process Flow - Worker processes and connections

Start nginx master process

↓

Spawn worker processes

↓

Each worker listens on ports

↓

Accept client connections

↓

Handle requests concurrently

↓

Close connections when done

↓

Wait for new connections

↩Loop back to Accept client connections

nginx master starts worker processes, each worker handles many client connections concurrently in a loop.

Execution Sample

Nginx

worker_processes 2;
worker_connections 1024;

# Each worker can handle 1024 connections
# Total max connections = 2 * 1024 = 2048

Configures nginx to run 2 worker processes, each able to handle 1024 simultaneous connections.

Process Table

Step	Worker Process	Connections Opened	Action	Total Connections
1	Worker 1	0	Start listening	0
2	Worker 2	0	Start listening	0
3	Worker 1	512	Accept connections	512
4	Worker 2	512	Accept connections	1024
5	Worker 1	1024	Max connections reached	1536
6	Worker 2	1024	Max connections reached	2048
7	Worker 1	1024	Close some connections	1924
8	Worker 2	900	Close some connections	1824
9	Worker 1	1024	Wait for new connections	1824
10	Worker 2	900	Wait for new connections	1824
11	All	-	No more connections to accept	1824

💡 No more incoming connections; workers wait idle.

Status Tracker

Variable	Start	After Step 3	After Step 4	After Step 5	After Step 6	After Step 8	Final
Worker 1 Connections	0	512	512	1024	1024	1024	1024
Worker 2 Connections	0	0	512	512	1024	900	900
Total Connections	0	512	1024	1536	2048	1824	1824

Key Moments - 3 Insights

Why does each worker process have its own connections instead of sharing one pool?

Each worker process runs independently to avoid locking and improve performance. See steps 3 and 4 where each worker accepts connections separately.

What happens when a worker reaches its max connections?

It stops accepting new connections until some close. See steps 5 and 6 where workers reach 1024 connections and stop accepting more.

Why do total connections sometimes decrease in the table?

Because some connections close after handling requests, freeing slots for new ones. See steps 7 and 8 where connections close and totals drop.

Visual Quiz - 3 Questions

Look at the execution table, how many connections does Worker 1 have after step 5?

A 900

B 512

C 1024

D 0

💡 Hint

Check the 'Worker 1 Connections' column at step 5 in the execution_table.

At which step do both workers reach their max connections?

A Step 4

B Step 6

C Step 5

D Step 7

💡 Hint

Look for when both workers have 1024 connections in the execution_table.

If worker_processes is set to 4 instead of 2, how does total max connections change?

A It doubles

B It stays the same

C It halves

D It quadruples

💡 Hint

Total max connections = worker_processes * worker_connections, see execution_sample.

Concept Snapshot

nginx uses worker_processes to run multiple workers.
Each worker handles up to worker_connections clients.
Total max connections = worker_processes * worker_connections.
Workers accept and close connections independently.
This improves performance and concurrency.

Full Transcript

In nginx, the master process starts multiple worker processes. Each worker listens for client connections independently. The number of worker processes is set by the 'worker_processes' directive. Each worker can handle a number of simultaneous connections defined by 'worker_connections'. Together, total max connections equal worker_processes times worker_connections. Workers accept connections, handle requests, then close connections and wait for new ones. This design avoids locking and improves performance by distributing load across workers.