Download as DOCX, PDF, TXT or read online from Scribd
Download as docx, pdf, or txt
You are on page 1/ 3
Docker
Docker is used to containerize the applications to solve system compatibility issues.
Docker creates images of the application so that it can run on
different OSs with different configurations.
How it does it is that it follows these steps: Flask, Flasgger
st 1. First of all, it creates a base image 1 (i.e., Linux: Ubuntu) Anaconda: Python 3._ from the docker hub. environment The base image talks about the basic OS we want. A base image in Docker is a crucial building block for creating Linux, Ubuntu Version containerized applications. It provides a reliable, efficient, and secure foundation on which you can layer your application Base Image code and configurations, making it easier to develop, deploy, and manage containerized applications across different environments.
Base images are essentially pre-built and pre-configured images of an operating
system or a specific software stack that provide a foundation for your own application or service.
Common base images include those based on popular Linux distributions like Ubuntu, CentOS, or Alpine Linux. Additionally, there are specialized base images for various programming languages, frameworks, and applications, which can save you even more time when containerizing specific types of software.
2. Next, we stack a component atop of base image where we specify the environment we will work with; in this case, it is Anaconda: Python 3.x.
3. Then we stack the component where we can mention all the required python packages; ones we have used in our app (like Flask, Flasgger, Pandas, NumPy)
4. Now, once we put together all the components we require, we will encapsulate the containers together and that will be the docker image.
These are the commands it will use: Docker Image
1. FROM: This is used to create the base image. Flask, Flasgger
Anaconda: Python 3._ Components
2. COPY: This is used to create the user root environment folder in docker. When we create an app in the host system Linux, Ubuntu Version Dockerized (the system that we create the app with), weContainer will have a folder structure where all the components of the app will be present. Base Image Using this ‘COPY’ we will be able to duplicate this folder structure from host system into the user root folder inside the docker container as well.
3. EXPOSE: Each and every docker image has a network interface. So, we expose the network interface of this docker image so that we can access the web app inside the docker image. The common network interface would be: 5000/8000.
Docker 1 4. WORKDIR: Now, if the working directory in the host system is like: Money:/…../…../ Where, ‘Money’ is the base folder in the host system, then we should also know in the docker container, from where we will initialize the app. We can take the working directory in the Docker as: user/python/local
5. RUN: We may need to install all the libraries that we used in the app in the host system in the docker image as well. Therefore, we create the requirements.txt file and mention all the libraries that we require. Then we use the command: pip install -r requirements.txt
6. CMD: If we look at the local host systems’ files of the app, there will be a file called: ‘app.py’. This will be the point of initialization for the entire app. To execute this file, we will use: ‘python app.py’ in the python terminal.
Similarly, in order to run the docker image, we will use the file(‘app.py’) inside ‘user/python/local’, and we will write the command: ‘python app.py’ there as well to execute the docker image. Here, the . represents the current working Steps: directory 1. Write the Docker file 2. Building the docker image -t is used to tell Docker to tag the docker 3. Running the app (money authenticator app) container in the name: money_api
Build the docker image:
-p is used to publish the container's port to docker build -t money_api . the host. Here, “-p 8000:8000” is specifying that port 8000 on the host should be See the dockers mapped to port 8000 on the container. docker ps -a Run the docker: docker run -p 8000:8000 money_api - This is the base image. It is popular because it installs some of the Create/write the docker file in the working library/package dependencies directory in the host system. It should have the automatically (like sklearn, etc.) following commands in it: FROM continuumio/anaconda3:4.4.0 - The . in COPY denotes working COPY . /usr/app/ directory of local host system. EXPOSE 8000 /usr/app denotes user root folder in WORKDIR /usr/app/ docker image. Hence, we are RUN pip install -r requirements.txt copying the files from host system CMD python app.py into docker image. - We are then exposing 5000 as the n/w interface
- Working directory will be the same
as in COPY’s second statement The requirements.txt (/user/app/) file will have the following libraries - Then we run the pip install -r listed along with their requirements.txt command to install versions: all the required libraries.
