Quick Answer

# Show live resource usage for all containers
docker stats

# Show only a specific container
docker stats my-container

# Grab a single snapshot instead of streaming
docker stats --no-stream

What You’re Trying to Do

A container feels slow, or the host’s CPU or memory seems maxed out, and you need to know exactly which container is responsible. docker stats shows per-container CPU usage, memory usage, network I/O, and disk I/O in real time — it’s the Docker equivalent of top or htop.

Environment

  • Docker Engine 24.x or later (Docker Desktop works the same way)
  • OS: Linux / macOS / Windows (WSL2)

Solution

Basic: Show Usage for All Containers

docker stats

Example output:

CONTAINER ID   NAME           CPU %     MEM USAGE / LIMIT     MEM %     NET I/O           BLOCK I/O
a1b2c3d4e5f6   web-app        2.34%     120.5MiB / 1.944GiB   6.05%     1.2MB / 850kB     0B / 0B
b2c3d4e5f6a1   db             15.67%    512.3MiB / 1.944GiB   25.72%    3.4MB / 2.1MB     45MB / 12MB

Press Ctrl + C to exit. By default the display refreshes every 1-2 seconds.

Show Only Specific Containers

docker stats web-app db

Pass multiple container names or IDs separated by spaces.

Grab a Single Snapshot (—no-stream)

Useful in scripts where you want one reading instead of continuous monitoring.

docker stats --no-stream

Customize the Output Columns (—format)

Use a Go template to control which fields are printed.

docker stats --format "table {{.Name}}\t{{.CPUPerc}}\t{{.MemUsage}}"

Example output:

NAME       CPU %     MEM USAGE / LIMIT
web-app    2.34%     120.5MiB / 1.944GiB
db         15.67%    512.3MiB / 1.944GiB

Output as JSON (for Scripting)

docker stats --no-stream --format "{{json .}}"

Combine with jq to extract specific fields.

docker stats --no-stream --format "{{json .}}" | jq -r '.Name + ": " + .CPUPerc'

Log Usage at a Regular Interval

while true; do
  docker stats --no-stream --format "{{.Name}},{{.CPUPerc}},{{.MemUsage}}" >> stats.csv
  sleep 60
done

Common Errors

Numbers Stay at 0.00% and Never Change

The container may be nearly idle, or you only checked without --no-stream for a split second. Put some load on it first.

# Generate load inside the container, then check
docker exec my-container stress --cpu 2 --timeout 30s
docker stats my-container

MEM LIMIT Looks Too High or Too Low

Unless you set an explicit limit with docker run --memory, the displayed limit is the host’s total available memory.

# Start a container with an explicit memory limit
docker run --memory="512m" --name my-container my-image

docker stats Feels Slow

With a large number of running containers, the live display can lag. Narrow the scope to speed it up.

docker stats $(docker ps --format "{{.Names}}" | grep app)

CPU Percentage Goes Above 100% on Windows/WSL2

This is expected when a container uses multiple CPU cores — CPU% is reported per core (e.g., up to 400% on a 4-core host), not capped at 100%.

FAQ

Q: What’s the difference between docker stats and docker top? docker stats shows aggregate CPU, memory, and I/O metrics for a container in real time. docker top lists the processes currently running inside a container — a different purpose entirely.

Q: Can I see stats for stopped containers? No. docker stats only reports on running containers; stopped containers don’t appear.

Q: Why can CPU usage exceed 100%? CPU% in docker stats is calculated against the host’s total logical cores, so fully utilizing multiple cores can push it well past 100% (e.g., up to 400% on 4 cores).

Q: Does this work with docker-compose projects? Yes. docker stats treats containers started by Docker Compose the same as any other container. Pass the specific container name if you only want one service.

Q: How do I actually cap resource usage instead of just watching it? Use --cpus and --memory when starting the container.

docker run --cpus="1.5" --memory="1g" my-image

Q: Is there an equivalent for Kubernetes? kubectl top pod shows similar per-pod CPU and memory usage (requires the metrics-server add-on).

Recommended VPS / Hosting

Build your production environment on a reliable VPS: