Welcome

This repo contains the code required to start and configure a local Octopus installation demonstrating common enterprise deployment patterns. It is based around Docker Compose and Terraform, and has been designed to allow the Octopus stack to be created and destroyed with ease.

Support Levels

The process implemented by the scripts in this repo are part of a pilot programme to support platform engineering teams. It incorporates tools with varying levels of official support.

The Octopus support teams will make reasonable endeavours to support teams that wish to use this process. However, existing support Service Level Agreements (SLAs) do not apply to the tools used by this repo, and should not be relied on for production deployments.

Getting Started

Start the Octopus and Git stack with the following command. Any missing tools or undefined environment variables will be reported before the setup can start:

./initdemo.sh

Shut the Octopus and Git stack down with:

./cleanup.sh

Common Prerequisites

Windows users should run this script in WSL.

You must have Docker and Docker Compose installed.

Linux users must download the Compose plugin. This page has instructions.

You also require these additional dependencies:

curl
Terraform
Minikube
Openssl
jq
argocd

Ubuntu dependencies are installed with the command:

curl --silent https://raw.githubusercontent.com/OctopusSolutionsEngineering/EnterprisePatternsReferenceImplementation/main/ubuntu-dependencies.sh | sudo bash

ARM macOS Prerequisites

Newer macOS users must enable the Use Rosetta for x86/amd64 emulation on Apple Silicon option in Docker:

Octonaught Prerequisites

Copy the contents of the shared note called Sample environment vars for EnterprisePatternsReferenceImplmenetation in the password manager to ~/.profile (for Linux) or ~/.zshrc (for macOS).

External User's Prerequisites

You must define the following environment variables:

OCTOPUS_SERVER_BASE64_LICENSE - set to a base 64 encoded Octopus license in the environment variable.
TF_VAR_docker_username - set to you DockerHub username.
TF_VAR_docker_password - set to your DockerHub password.

Typically, this is done by adding the following like to ~/.profile (for Linux) or ~/.zshrc (for macOS):

export OCTOPUS_SERVER_BASE64_LICENSE=PExpY2Vuc2UgU2lnbmF0dXJlPSJk...
export TF_VAR_docker_username=your_dockerhub_username
export TF_VAR_docker_password=your_dockerhub_password

Optional Settings

Set these environment variables if you want the Slack incident channel runbook to work:

TF_VAR_slack_bot_token - set to a Slack bot token (i.e. a token starting with xoxb-).
TF_VAR_slack_support_users - set to the comma separated list of Slack user IDs that will be pulled into the incident channels.

Set these variables to complete the deployment of the sample Azure application:

TF_VAR_azure_application_id - set to your Azure application ID.
TF_VAR_azure_subscription_id - set to your Azure subscription ID.
TF_VAR_azure_password - set to your Azure password.
TF_VAR_azure_tenant_id - set to your Azure tenant ID.
INSTALL_ARGO - set to TRUE to install and populate ArgoCD in minikube.

FAQ

Q. How do I fix the "command failed" err="failed complete: too many open files" error some minikube pods have?

A. This post has some suggestions. Linux users can run these commands:

sudo sysctl fs.inotify.max_user_instances=8192
sudo sysctl fs.inotify.max_user_watches=524288
sudo sysctl -p

This page has details on setting these values permanently.

Q. Running KUBECONFIG=/tmp/octoconfig.yml minikube tunnel in WSL says sudo permissions will be asked for it. Where is the sudo prompt?

A. It appears that you can ignore these messages. ArgoCD may take a while to start (so you may need to refresh the page a few times), but will eventually be available on the EXTERNAL-IP shown by the command KUBECONFIG=/tmp/octoconfig.yml kubectl get serviceargocd-server -n argocd.

ryanrousseau/EnterprisePatternsReferenceImplementation