Deployment Environment Setup

1. Install the Google Cloud SDK

2. Set the current project as reach4help:

   gcloud config set project reach4help
   

3. Login:

   gcloud auth login
   

4. [OPTIONAL] Using the gcloud command line tool, install the Kubernetes command-line tool. kubectl is used to communicate with Kubernetes, which is the cluster orchestration system of GKE clusters:

   gcloud components install kubectl
   

Deployment

1. Run the following command in the root of the repo, which will submit the repo to cloud build, and then once the build completes the image will be stored in the container repository:

   gcloud builds submit --tag gcr.io/reach4help/reach4help-chatbot
   

2. Deploy the latest version of the image (can be done in the Google Cloud Shell)

   1. Get the bot credentials

      gcloud container clusters get-credentials reach4help-bot --zone=us-central1-b
      

   2. Tell kubernetes to rollout the image to the bot

      kubectl set image deployment reach4help-bot reach4help-chatbot=gcr.io/reach4help/reach4help-chatbot@sha256:<sha from deployment logs>
      

3. Check that the bot successfully restarted and posted a message in the #bot-admin slack channel.

Redeploying deployment.yaml config

If you have to change deployment.yaml for any reason, such as moving an environment variable into a secret, then updating the image is not enough and you need to apply the changed config like so:

kubectl apply -f deployment.yaml

Updating Secrets

If you need to upload new secrets to kubernetes, create them in a file called .env.secrets (which is ignored by gitignore), check that the bot works locally, and then upload the secrets with the following command:

kubectl create secret generic chatbot --from-env-file ./.env.secrets --dry-run -o yaml | kubectl apply -f -

At which point redeploying the app should allow the new environment variables to take affect.