Container resource usage stats in Docker - Step-by-Step Execution

Process Flow - Container resource usage stats

Start docker stats command

↓

Docker daemon collects stats

↓

Retrieve CPU, Memory, Network, I/O data

↓

Display stats in real-time

↓

User views resource usage

↓

Stop command to end stats

The docker stats command starts, the daemon collects container resource data, then displays it live until stopped.

Execution Sample

Docker

docker stats --no-stream
CONTAINER ID   NAME       CPU %   MEM USAGE / LIMIT   MEM %   NET I/O     BLOCK I/O   PIDS
abc123def456   my_app     2.34%   50MiB / 1GiB       4.88%   1.2MB / 0.8MB  10MB / 5MB  12

This command shows a snapshot of resource usage for running containers without continuous updates.

Process Table

Step	Action	Data Collected	Output Displayed
1	Run 'docker stats --no-stream'	Start stats collection	Header row with column names
2	Docker daemon collects CPU usage	CPU % for each container	CPU % column shows e.g. 2.34%
3	Docker daemon collects Memory usage	Memory used and limit	MEM USAGE / LIMIT and MEM % columns filled
4	Docker daemon collects Network I/O	Bytes sent and received	NET I/O column shows e.g. 1.2MB / 0.8MB
5	Docker daemon collects Block I/O	Disk read/write bytes	BLOCK I/O column shows e.g. 10MB / 5MB
6	Docker daemon collects PIDs	Number of processes	PIDS column shows e.g. 12
7	Display all collected data in table	All stats combined	Full stats row for container displayed
8	Command ends (no streaming)	Stop collecting	Stats output ends after one snapshot

💡 Command ends after one snapshot because of --no-stream option

Status Tracker

Variable	Start	After Step 2	After Step 3	After Step 4	After Step 5	After Step 6	Final
CPU %	N/A	2.34%	2.34%	2.34%	2.34%	2.34%	2.34%
Memory Usage	N/A	N/A	50MiB / 1GiB	50MiB / 1GiB	50MiB / 1GiB	50MiB / 1GiB	50MiB / 1GiB
Memory %	N/A	N/A	4.88%	4.88%	4.88%	4.88%	4.88%
Network I/O	N/A	N/A	N/A	1.2MB / 0.8MB	1.2MB / 0.8MB	1.2MB / 0.8MB	1.2MB / 0.8MB
Block I/O	N/A	N/A	N/A	N/A	10MB / 5MB	10MB / 5MB	10MB / 5MB
PIDs	N/A	N/A	N/A	N/A	N/A	12	12

Key Moments - 3 Insights

Why does the 'docker stats' command keep updating the output by default?

By default, 'docker stats' streams live data continuously. In the execution_table, step 8 shows the command ends only because of the '--no-stream' option.

What does the 'MEM USAGE / LIMIT' column represent?

It shows how much memory the container uses versus its maximum allowed memory. Step 3 in the execution_table shows this data is collected and displayed.

Why might CPU % be low even if the container is busy?

CPU % is a snapshot of usage at that moment. If the container is busy intermittently, the snapshot might catch it at a low usage time, as shown in step 2.

Visual Quiz - 3 Questions

Look at the execution_table at step 4, what resource data is collected?

A Memory usage

B Network I/O

C CPU usage

D Block I/O

💡 Hint

Check the 'Data Collected' column at step 4 in execution_table

At which step does the command stop collecting stats due to the '--no-stream' option?

A Step 6

B Step 7

C Step 8

D Step 5

💡 Hint

Look at the 'exit_note' and step 8 in execution_table

According to variable_tracker, what is the value of 'PIDs' after step 6?

A N/A

B 12

C 0

D Unknown

💡 Hint

Check the 'PIDs' row in variable_tracker under 'After Step 6'

Concept Snapshot

docker stats command shows live resource usage of containers
By default, it streams continuously; '--no-stream' shows one snapshot
Displays CPU %, memory usage and limit, network I/O, block I/O, and PIDs
Useful for monitoring container health and performance
Stops when user interrupts or with '--no-stream' option

Full Transcript

The docker stats command starts by sending a request to the Docker daemon to collect resource usage data for running containers. The daemon gathers CPU usage, memory usage and limits, network input/output, block device input/output, and the number of processes (PIDs) for each container. This data is displayed in a table format with columns for each metric. By default, docker stats streams this data live, updating continuously until the user stops it. Using the '--no-stream' option makes the command show only one snapshot and then exit. This helps users monitor container resource consumption in real time or as a single report.