binhbt/Kafka-flask-docker-demo

Kafka Demo

Notes 2019-08-21: Updated for Python 3.7 and kafka-python 1.4.6

This repo was written as part of a job interview and isn't necessarily a good example of a production Kafka use case. The requirements imposed by the interviewer and the limited time I had to spend on it meant that this was stripped down to something very limited. But it is a useful reference point and contains a decent demo of Kafka's pub/sub capabilities.

Inside you'll find a small handful of goodies.

• docker-compose.yml
  • This is the base composition. It should allow you to bring up the stack with just docker-compose up
• docker-compose.dev.yml
  • The same as docker-compose.yml except that it mounts the producer and consumer src directories to make the dev/test cycle easier. You can start it with docker-compose -f docker-compose.dev.yml up
• producer
  • The source and Dockerfile for the producer application.
• consumer
  • Same thing but for the consumer
• tester
  • Some tests to exercise the stack once it's up and running. You might expect these to be unit tests but they're actually integration tests. The applications are very simple so the meat of the testing was really in the component interactions.

Dependencies

Other versions may work but this build has been tested with the following:

Python 3.7.x docker-compose version 1.25.0-rc1 Docker version 19.03.1

You should also run:

pip install -r requirements.txt

in the tester directory to make sure you have the packages needed if you plan to run the test scripts.

Example Usage

You'll always need to start it manually

docker-compose up

It'll take a few seconds to come up. Usually the producer and consumer seem to start before kafka is ready and there's a point where it looks like it's good to go, but it's not.

consumer_1   | Running app in debug mode!
consumer_1   |  * Running on http://0.0.0.0:80/ (Press CTRL+C to quit)
consumer_1   |  * Restarting with stat
producer_1   |  * Running on http://0.0.0.0:80/ (Press CTRL+C to quit)
producer_1   |  * Restarting with stat
consumer_1   |  * Debugger is active!
consumer_1   |  * Debugger PIN: 325-703-385
producer_1   |  * Debugger is active!
producer_1   |  * Debugger PIN: 111-020-208

You shoud wait a few more seconds until you see this message in the console:

kafka_1      | creating topics: stats:1:1

Manual Testing

In another terminal you can then run:

curl -X POST -H "Content-Type: application/json" localhost:8282/events -d '{"event" : "a test event"}'

output

{"status": "success"}

Then, to read the message from the queue:

curl localhost:8283/events

Scripted Testing

After starting the stack with docker-compose go into the tester directory and run:

python tests.py

output

.....
----------------------------------------------------------------------
Ran 5 tests in 0.668s

OK

Kafkacat

If you want to test using kafkacat you'll need to jump through some minor hoops. The issue is that Kafka advertises it's broker through Zookeeper using it's hostname and in the docker composition the Kafka host is 'kafka'. But you'll likely be trying to run kafkacat locally against the exported 'localhost' port, which can't resolve the advertised hostname of 'kafka'. To fix this just add an entry to your /etc/hosts file (or equivalent for windows). eg:

127.0.0.1       localhost kafka

Dev Mode

A quick note about starting with the docker-compose.dev.yml file. This one will mount the local directories to the containers to make dev easier, but if you've already started the containers with the docker-compose.yml file there will be a conflict.

Make sure you run docker-compose down before switching to dev mode.

Known Issues and Other things to worry about

This is just a quick demo of kafka's pub sub using Python. It's missing a lot of stuff you would want if you were going to use this in a production environment. Including:

• payload validation
• authentication and authorization
• improved error handling
• connection retries for kafka brokers
• managing partitions
• probably other things