Environment notice on Kubernetes

A recent project of mine has been getting a Rails application set up on Kuberentes.

I’ve found myself frequently opening up shells and Rails consoles on the pods for debugging and testing. It didn’t take long before I almost ran something in production that I meant to run in the staging environment.

Sometimes a simple alert is enough to keep you from shooting yourself in the foot.

I created a simple Bash script, added it to the Docker image, and added it to both the Bash and irb profiles so that it would be executed when starting either a Bash shell or a Rails console and let me know which environment I’m in.


In our setup, Kubernetes namespaces denote the environment such as production or staging.

On a running pod, you can read the current Kuberentes namespace from a file mounted by Kubernetes:

$ cat /var/run/secrets/kubernetes.io/serviceaccount/namespace

As a fallback, try reading the environment variable KUBERNETES_NAMESPACE which we set globally. This could come in handy if the namespace file path ever changes.


set -euo pipefail



if [ -f ${kse_namespace_file} ]; then
  namespace=$(cat ${kse_namespace_file})

case ${namespace} in
  production) style="\e[91m\e[21m\e[7m" ;;
  staging)    style="\e[93m" ;;
  edge)       style="\e[36m" ;;
  *)          style="\e[95m" ;;

echo -e "  This is the ${style}${namespace}\e[0m environment."


In the Dockerfile, env_alert.sh is added to /opt/bin.

To run it when starting an interactive Bash shell, append to ~/.bashrc

To run it when starting a Rails console, add a line of Ruby to ~/.irbrc

COPY docker/env_alert.sh /opt/bin/env_alert

RUN echo /opt/bin/env_alert >> ~/.bashrc
RUN echo 'system "/opt/bin/env_alert"' >> ~/.irbrc


Now the script runs every time a shell or Rails console is started.

Big warning in the production environment:

More subtle notice in the edge environment: