How to fix Cannot connect to the Docker daemon at tcp://localhost:2375. Is the docker daemon running? in Docker

First, verify if DOCKER_HOST is explicitly set:

echo $DOCKER_HOST

If this returns tcp://localhost:2375 or similar, that's why Docker is trying to connect via TCP.

To unset it temporarily:

unset DOCKER_HOST
docker version

If Docker works after unsetting, the fix is to remove DOCKER_HOST from your shell configuration files (~/.bashrc, ~/.zshrc, ~/.profile).

If you're using WSL2 with Docker Desktop, you should NOT use the TCP connection method. Docker Desktop provides native WSL2 integration:

1. Open Docker Desktop
2. Go to Settings (gear icon)
3. Navigate to Resources > WSL Integration
4. Enable integration for your WSL distribution
5. Click Apply & Restart

Then in your WSL terminal:

# Remove DOCKER_HOST from your shell config
sed -i '/DOCKER_HOST/d' ~/.bashrc
sed -i '/DOCKER_HOST/d' ~/.zshrc

# Reload your shell
source ~/.bashrc

# Verify Docker works
docker version

With WSL2 integration, Docker commands work natively without any environment variables.

If you specifically need TCP access (WSL1, or accessing Docker from another application):

1. Open Docker Desktop
2. Go to Settings (gear icon)
3. Check "Expose daemon on tcp://localhost:2375 without TLS"
4. Click Apply & Restart

Warning: This exposes Docker without encryption or authentication. Only use this on trusted networks, and never on publicly accessible machines.

Then verify the connection:

export DOCKER_HOST=tcp://localhost:2375
docker version

On Linux, you need to configure the Docker daemon to listen on TCP:

Method 1: Using systemd override (recommended)

Create a systemd drop-in file:

sudo mkdir -p /etc/systemd/system/docker.service.d
sudo nano /etc/systemd/system/docker.service.d/override.conf

Add the following content:

[Service]
ExecStart=
ExecStart=/usr/bin/dockerd -H fd:// -H tcp://127.0.0.1:2375

Then reload and restart Docker:

sudo systemctl daemon-reload
sudo systemctl restart docker

Method 2: Using daemon.json

Edit /etc/docker/daemon.json:

{
  "hosts": ["unix:///var/run/docker.sock", "tcp://127.0.0.1:2375"]
}

Then restart Docker:

sudo systemctl restart docker

Security Note: Binding to 127.0.0.1 restricts access to localhost only. Never bind to 0.0.0.0 without TLS and authentication.

In GitLab CI with Docker-in-Docker (dind), the Docker daemon runs as a separate service, not on localhost:

# .gitlab-ci.yml
image: docker:latest

services:
  - docker:dind

variables:
  # Use 'docker' hostname, not 'localhost'
  DOCKER_HOST: tcp://docker:2375
  # For Docker 19.03+ with TLS:
  # DOCKER_HOST: tcp://docker:2376
  # DOCKER_TLS_CERTDIR: "/certs"

build:
  script:
    - docker version
    - docker build -t myimage .

Important: The hostname docker refers to the dind service container. Using localhost or 127.0.0.1 will fail because the daemon is in a different container.

Different CI/CD systems have different configurations for Docker-in-Docker:

Jenkins:

pipeline {
  agent {
    docker {
      image 'docker:dind'
      args '--privileged'
    }
  }
  environment {
    DOCKER_HOST = 'tcp://localhost:2375'
  }
}

GitHub Actions:

jobs:
  build:
    runs-on: ubuntu-latest
    services:
      docker:
        image: docker:dind
        options: --privileged
    steps:
      - name: Build
        env:
          DOCKER_HOST: tcp://docker:2375
        run: docker build .

Bitbucket Pipelines:
Enable Docker in your pipeline configuration and use the default socket, or configure TCP access on the runner.

To debug further, check if Docker is running and listening on the expected port:

Check if Docker daemon is running:

# On Linux
sudo systemctl status docker

# On Windows (PowerShell)
Get-Service *docker*

Check if port 2375 is listening:

# On Linux
ss -tlnp | grep 2375
# or
netstat -tlnp | grep 2375

# On Windows (PowerShell)
netstat -an | findstr 2375

Test TCP connection:

curl http://localhost:2375/version
# or
nc -zv localhost 2375

If nothing is listening on port 2375, the daemon needs to be configured to listen on TCP (see previous steps).

If Docker fails to bind to port 2375, another application might be using it:

Find what's using port 2375:

# On Linux
sudo lsof -i :2375
# or
sudo ss -tlnp | grep 2375

# On Windows (PowerShell as Administrator)
netstat -ano | findstr 2375
# Then find the process
Get-Process -Id <PID>

Common conflicts:
- Another Docker daemon instance
- Docker Desktop already running
- Kubernetes or minikube
- Development tools that proxy Docker

Stop the conflicting application or configure Docker to use a different port.