Skip to content

Latest commit

 

History

History
233 lines (155 loc) · 5.56 KB

README.md

File metadata and controls

233 lines (155 loc) · 5.56 KB

This is an example of how to create a Python Flask web server, run it in an Alpine Linux docker container and deploy that container to an EC2 instance on AWS using terraform. This follows from DDoS-attack-detection-using-HTTP-packet-clustering-pattern and thanks to gurpreet0610 for sharing that and In terraform, how can I pull a Docker image on the target? for suggeting it.

Python Flask Server

From the command line

Create a test environment

Use the python3 venv module to setup a virtual test environment.

Create the venv directory

This is a one time setup

python3 -m venv venv-alpfs

Activate

This is needed each time a new shell is started for testing

source venv-alpfs/bin/activate

Install

One time setup again

pip install -r requirements.txt

In the activated test environment

./alpfs.py -l debug
 * Serving Flask app 'alpfs' (lazy loading)
 * Environment: production
   WARNING: This is a development server. Do not use it in a production deployment.
   Use a production WSGI server instead.
 * Debug mode: off
[2021-05-31 07:40:37,558 WARNING _internal.py 225]  * Running on all addresses.
   WARNING: This is a development server. Do not use it in a production deployment.
[2021-05-31 07:40:37,558 INFO _internal.py 225]  * Running on http://192.168.4.26:80/ (Press CTRL+C to quit)


^C

To test

With a browser

The web page should now be available on http://localhost

With client.py

client.py is a python script that uses the python requests module to access the API.

./client.py -l debug
[2021-06-06 07:04:08,525 DEBUG client.py 57] GET http://localhost:80/api/v0/whoami headers: None, params: {'foo': 'bar', 'baz': 42}
[2021-06-06 07:04:08,536 DEBUG connectionpool.py 227] Starting new HTTP connection (1): localhost:80
[2021-06-06 07:04:08,538 DEBUG connectionpool.py 452] http://localhost:80 "GET /api/v0/whoami?foo=bar&baz=42 HTTP/1.1" 200 73
<Response [200]>
[2021-06-06 07:04:08,539 DEBUG client.py 76] POST http://localhost:80/api/v0/whoareyou headers: None, JSON:{'foo': 'bar', 'baz': 42},  files:None
[2021-06-06 07:04:08,540 DEBUG connectionpool.py 227] Starting new HTTP connection (1): localhost:80
[2021-06-06 07:04:08,542 DEBUG connectionpool.py 452] http://localhost:80 "POST /api/v0/whoareyou HTTP/1.1" 200 71
<Response [200]>

With client.js

client.js is a node.js sketch of a script that uses the axios module to access the API. You will need to run npm install.

node client.js -u whoami
action response { args: {}, timestamp: '2021-06-06T09:02:51-0700' }

With Python unittest

test_alpfs.py is an example using the built in python unittest module with flask's built in test client and does not need to have the server running in another process.

python3 test_alpfs.py -v
test_get_whoami (__main__.AlpfsTests)
Test GET whoami API ... ok
test_get_whoareyou (__main__.AlpfsTests)
Test GET whoareyou API fails ... ok
test_home_page (__main__.AlpfsTests)
Test home page ... ok
test_post_whoami (__main__.AlpfsTests)
Test POST whoami API fails ... ok
test_post_whoareyou (__main__.AlpfsTests)
Test POST whoareyou API ... ok
test_starbug_link (__main__.AlpfsTests)
Test home page has starbug.com link ... ok

----------------------------------------------------------------------
Ran 6 tests in 0.034s

OK

Alpine Linux container

To build

docker build -f Dockerfile -t alpfs .

To run

docker run --name alpfs00 -d -p 80:80 alpfs

The web page should now be available on http://localhost

With port 8080 on the outside mapped to port 80 on the inside

docker run --name alpfs00 -d -p 8080:80 alpfs

The web page should now be available on http://localhost:8080

Logs

docker logs alpfs00
* Serving Flask app 'alpfs' (lazy loading)
 * Environment: production
   WARNING: This is a development server. Do not use it in a production deployment.
   Use a production WSGI server instead.
 * Debug mode: off
[2021-05-31 14:49:25,951 WARNING _internal.py 225]  * Running on all addresses.
   WARNING: This is a development server. Do not use it in a production deployment.
[2021-05-31 14:49:25,952 INFO _internal.py 225]  * Running on http://172.17.0.2:80/ (Press CTRL+C to quit)
[2021-05-31 14:49:37,880 INFO _internal.py 225] 172.17.0.1 - - [31/May/2021 14:49:37] "GET / HTTP/1.1" 200 -
[2021-05-31 14:49:37,885 INFO _internal.py 225] 172.17.0.1 - - [31/May/2021 14:49:37] "GET /static/style.css HTTP/1.1" 200 -

Shell

Use docker exec to open a shell in the container

docker exec -it alpfs00 sh

To stop

docker stop alpfs00
docker ps -a
CONTAINER ID   IMAGE     COMMAND             CREATED              STATUS                       PORTS     NAMES
4599c0076e3a   alpfs     "python alpfs.py"   About a minute ago   Exited (137) 6 seconds ago             alpfs00

Tag and push to DockerHub

docker tag alpfs lrmcfarland/alpfs
docker push lrmcfarland/alpfs

The tag is used by terraform to pull this image from DockerHub.

Terraform

see terraform/README.md