Sidharth R

Container Cpu Limit Configuration

Updated: ⚬ Page views: loading...

Introduction

In this lab, you’ll limit the CPU usage of a Docker container that consumes excessive CPU resources.

Scenario

A container named cpu_test is running from the myapp:cpu image and performs CPU-intensive computations. When monitored with docker stats, the container regularly consumes 150–200% CPU, utilizing multiple CPU cores and impacting other services on the host.

Task

Recreate the container with a CPU limit of 500m CPU so that docker stats reports CPU usage consistently below 60% during validation.

Solution

Let’s check the container’s current CPU usage.

Container Stats
$ docker stats

CONTAINER ID   NAME       CPU %     MEM USAGE / LIMIT     MEM %     NET I/O       BLOCK I/O   PIDS
dec7403ffc96   cpu_test   199.78%   6.453MiB / 1.042GiB   0.60%     946B / 126B   0B / 0B     4

Stop and remove the existing container.

BASH
docker stop dec
docker rm dec

Check current CPU limits:

BASH
docker inspect cpu_test | grep -i cpu
CPU Configuration
"Path": "/app/cpu_workload",
"Name": "/cpu_test",
"CpuShares": 0,
"NanoCpus": 0,
"CpuPeriod": 0,
"CpuQuota": 0,
"CpuRealtimePeriod": 0,
"CpuRealtimeRuntime": 0,
"CpusetCpus": "",
"CpusetMems": "",
"CpuCount": 0,
"CpuPercent": 0,
"/app/cpu_workload"
"Image": "myapp:cpu",

Since all CPU-related settings are 0, no CPU limits are applied.

Recreate the container with a limit of 0.5 CPU (500m).

BASH
docker run -d \
  --name cpu_test \
  --cpus 0.5 \
  myapp:cpu

Verify the new CPU limit.

Container Stats
$ docker stats

CONTAINER ID   NAME       CPU %    MEM USAGE / LIMIT     MEM %    NET I/O      BLOCK I/O   PIDS
8640f9d896f7   cpu_test   50.06%   6.473MiB / 1.041GiB   0.61%    876B / 126B  0B / 0B     4

The container now uses approximately 50% CPU, confirming that the --cpus 0.5 limit has been applied successfully.

Keep reading