• A compatible cluster (e.g. Minikube)
• Kubernetes (1.16 and latter)
• Istio

To set up Twittify on a cluster, run ./setup.py --istio cloud. It verifies Istio installation and applies Twittify configs located at istio to the cluster.

To start a Minikube cluster, run ./setup.py --istio local. User must be in docker group.

To discover address of the cluster, run ./setup.py --address to show address of the Twittify client.

If cluster is running on Minikube, run ./setup.py --tunnel to show address and starts Minikube tunnel.

• Start up a cluster.
• Attach to the cluster via terminal.
• In terminal run

$ git clone https://github.com/team-cringe/twittify
$ cd twittify
$ ./setup.py --istio cloud

• If deployment was successful, services would be listed in Services & Ingress. By default, scraper must fetch tweets of at least 2500 users for clusterizer to start working. Considering that, the deployment may take quite a long time (~4 hours).
• In order to find out an address of the client, run ./setup.py --address.
• The service should be available when the clusterizer is ready.

Visit 35.228.148.166/client to see how the service works.

• pipenv
• docker

• Run ./setup.py --test.
• For each container, install dependencies using pipenv install.
• Launch main.py in corresponding folders.

Additionally, setup.py deploys Kibana container. By default, data is available in a browser at localhost:5601