Skip to content

Latest commit

 

History

History
31 lines (20 loc) · 1.95 KB

README.md

File metadata and controls

31 lines (20 loc) · 1.95 KB
Raw

Area-51

Deployment from scratch

Prerequisites

You'll need:

  1. ansible
  2. kubectl
  3. k3sup

Initial deployment

Before executing these steps, ensure you have an SSH keypair named Area 51 on openstack and you have the private key at ~/.ssh/area51. You will also need to cp openrc.example .openrc, and then fill in your openstack credentials in .openrc. Then:

  1. sudo chmod +x deploy/deploy-k3s-cluster.sh
  2. ./deploy/deploy-k3s-cluster.sh

Twitter-App

Few things to note of importance.

  1. The file twitter_streaming.py is the main file for executing the app for streaming from the Dockerfile.
  2. There are changes still to be made to this file in terms of what data we are to collect from each tweet. We will need to discuss this and agree on what this should look like.
  3. The file twitter_streaming_into_database.py is the original version of twitter_streaming.py except that it feeds tweets directly into a CouchDB on my local machine. This was mainly for testing as we now need to get the networking setup with the Docker containers for each harvester so as to feed into the database (I don't know how to do this yet).
  4. The environment variables (authentication keys) are missing from the Dockerfile. We need to add these to each container for each of the harvesters. This can be done in the Dockerfile I'm pretty sure using a .env file to accompany the Dockerfile but I haven't figured this part out just yet.
  5. I need to put together a docker-compose.yml file still.

To the run the app locally: docker run --name twitter-streamer -p 5000:5000 -d twitter-streamer

I still need to review/test some of the above but any help on linking the couchDB setup with the above would be great. I have it working locally but not in a clustered setup.