Elastic beans talk is a suitable service for distributing a application or docker container.
In our teams project, there is a some use case running a task periodically. At first, we choose AWS lambda and Cloudwatch to execute these use case but AWS lambda’s timeout limitation is fixed in 30 seconds.
So, we choose Elastic beans talk for alternatives. It can handle more time for running a task like 500 seconds.
Here be a tutorial post for deploying pure python application to Elastic beans talk.
1. Creating a Elastic beans talk application
2. Creating a environment
Choose a **’worker tier’**
3. Set a base configuration
Choose a platform. In this case, i’m using pure python(Not a Django, Flask).
4. Writing a script for deployment
For deploying a application, it needs a kind of option script. First, you shuold make a directory `.ebextensions` into python project root directory then write a option script named `options.config`. This configuration file is for designating where the python file is and set a number of application’s processes and threads.
* options.config
option_settings:
aws:elasticbeanstalk:container:python:
WSGIPath: application.py
NumProcesses: 3
NumThreads: 205. Scheduling a task
To run a task periodically, write a crontab script in `cron.yaml` file. If you confused with setting a time, [this site will be helpful](https://crontab.guru/)
* cron.yaml
version: 1
cron:
— name: “someTaskRun”
url: “/someTask”
schedule: “0 18 * * *”
7. Wrapping a python code
Wrapping a task file with python application. It will be executed by WSGI with url.
This code refers to the sample pure python code of Elastic beans talk.
import logging
import logging.handlersfrom wsgiref.simple_server import make_server# Create logger
logger = logging.getLogger(__name__)
logger.setLevel(logging.INFO)def application(environ, start_response):
path = environ[‘PATH_INFO’]
method = environ[‘REQUEST_METHOD’]logger.info(‘start ‘ + path)
response = ‘’if method == ‘POST’:
try:
if path == ‘/someTask’:
# do some task
passresponse = ‘succefully’ + path + ‘ run ‘except (TypeError, ValueError):
logger.warning(‘Error retrieving request body for async work.’)status = ‘200 OK’
headers = [(‘Content-type’, ‘text/html’)]start_response(status, headers)
return [response]if __name__ == ‘__main__’:
httpd = make_server(‘’, 8000, application)
print(“Serving on port 8000…”)
httpd.serve_forever()
