docker_deployment_bots.md

Bot Docker Setup

To run the bot in a containerized environment the repository provides a simple Dockerfile. It is possible to spin up containers one by one by using this docker file and chaning the .env file, however, it is also possible to spin up multiple instances of the bot using docker-compose mechanic from docker and the provided docker-compose.yml file (advised). Both of the options use the Dockerfile from the repository.
First off, make sure that Docker is correctly installed:

• Install Docker

The Dockerfile currently holds commands to start the index.ts file and install the required packages:

FROM node:19

ENV TERM=xterm
RUN mkdir /usr/src/app
WORKDIR /usr/src/app

COPY package*.json ./
RUN npm install

COPY tsconfig.json ./
COPY src ./src
RUN npm run build

CMD ["node", "out/index.js"]

Docker Run method

This specific section describes how to run a single instance of the bot in a containerized environment. It utilizes the Dockerfile and requires a few manual operations. If you plan on spinning up multiple instance of the bot, it might be easier to use the docker-compose method.

Creating Image

The first step in running the bot in a Docker Container is to create the image, to do so, run in the root folder of the repository:

docker build . -t [name you want]

Make sure to replace [name you want] with whatever name you would like for the image. {% hint style="info" %} Note: the images will use the environment variable described in .env, you have to change these before creating the image. {% endhint %}

When the image is succesfully created you can view it with:

docker images

#results in:
REPOSITORY        TAG       IMAGE ID       CREATED       SIZE
skipmev/juno      latest    f03ce8ffe244   5 weeks ago   1.13GB

Make sure the image you just created is listed in the above result of the docker image command.

Starting the Container

To start the previously created image in a Docker Container:

docker run -restart=always [name of your image]

Again, make sure to replace [name of your image] with the name of the previously created image. The --restart=always flag will make sure the bot will restart on an error. For example when we encounter a timeout by a RPC_ENDPOINT, the bot will restart and continue running. In our case the command would be:

docker run --restart=always skipmev/juno

The output will be the same as starting the bot with npm start. After executing the run command you will continue to see the logs in console, which you can safely close without the bot shutting down. To view all running containers:

docker ps -a

Docker Compose method

If you plan on running multiple instances of the bot that use different .env files, it might be more suitable to use the docker-compose method. This is supported by the provided docker-compose.yml file in the repository that has to be altered to your needs. For example when spinning up two containers, one for terra and one for juno, using the terra.env and the juno.env respectively, the file looks like this:

version: '3.9'

services:
 bot-terra:
   build: .
   image: white-whale-bot-terra
   container_name: white-whale-bot-terra
   env_file:
     - terra.env
   restart: always
   stdin_open: true
   tty: true
 
 bot-juno:
   build: .
   image: white-whale-bot-juno
   container_name: white-whale-bot-juno
   env_file:
     - juno.env
   restart: always
   stdin_open: true
   tty: true
   

Now to build an run both instances (services) we simply hit:

docker-compose up

and the containers should be running on completion, to check, hit:

docker ps -a