Run a Discord Bot 24/7 on a VPS with Docker

To run a Discord bot 24/7 on a VPS, you package the bot in a Docker image, pass the token as an environment variable and set a restart policy so Docker brings it back after a crash or a reboot. The steps below cover the Dockerfile, the env file, the run command and how to read the logs when something goes wrong.

Before you start

You need:

A Linux VPS with SSH access. Your IP, username and password are in your VPS provider's control panel.
Docker installed. If it is not there yet, follow Install Docker and Docker Compose on a Linux VPS.
A Discord application and bot user created in the Discord Developer Portal, with the bot token copied somewhere safe.
Either your own bot code or the small example below.
About 15 minutes.

Note: A bot that serves a few servers runs fine on a small VPS. A Python or Node bot usually idles well under 100 MB of RAM, so memory is rarely the limit — long log files are a more common problem.

Prepare the bot project

Create a working directory and move into it:

Bash

mkdir -p /opt/discord-bot
cd /opt/discord-bot

Create requirements.txt with the library:

discord.py

Then create bot.py with a minimal bot that replies to a ping:

Python

import os
import discord

TOKEN = os.environ["DISCORD_TOKEN"]

intents = discord.Intents.default()
intents.message_content = True
client = discord.Client(intents=intents)


@client.event
async def on_ready():
    print(f"Logged in as {client.user}", flush=True)


@client.event
async def on_message(message):
    if message.author == client.user:
        return
    if message.content == "!ping":
        await message.channel.send("pong")


client.run(TOKEN)

The token is read from the environment with os.environ, never hard-coded.

Tip: Once the bot works, pin the library version in requirements.txt so a rebuild months later does not pull a breaking change. Check the library's own docs for the version you should pin.

Write the Dockerfile

Create /opt/discord-bot/Dockerfile:

Dockerfile

FROM python:3-slim

ENV PYTHONUNBUFFERED=1
WORKDIR /app

COPY requirements.txt ./
RUN pip install --no-cache-dir -r requirements.txt

COPY bot.py ./

RUN useradd --create-home --shell /usr/sbin/nologin bot
USER bot

CMD ["python", "bot.py"]

Three things matter here. PYTHONUNBUFFERED=1 makes print output appear in docker logs immediately instead of sitting in a buffer. Copying requirements.txt before the code keeps the dependency layer cached, so code edits rebuild in seconds. And the USER bot line means the process does not run as root inside the container.

Build the image:

Bash

docker build -t my-discord-bot:1.0 .

Expect a few lines of pip output and a final line naming the image my-discord-bot:1.0.

Pass the token as an environment variable

Warning: Anyone who has your bot token controls your bot. Never put it in the Dockerfile, in bot.py, in a Git repository or on the command line — command-line values end up in the shell history.

Create the env file and lock it down:

Bash

nano /opt/discord-bot/bot.env
chmod 600 /opt/discord-bot/bot.env

Put one line inside it, with no quotes and no spaces around the =:

INI

DISCORD_TOKEN=PASTE_YOUR_TOKEN_HERE

Docker reads this file at start-up and injects DISCORD_TOKEN as an environment variable inside the container. The file stays on the host with mode 600, so only root can read it.