JourneyMonitor

About this repository

This repository is responsible for managing setups where JourneyMonitor applications can be run.

About the JourneyMonitor project

Please see ABOUT.md for more information.

Setup instructions

Setting up a development environment

The officially supported way to run a development environment is to use Docker and Docker Compose. It has been verified to work on Mac OS X with Docker Machine and Docker for Mac.

• Install Docker
• Install Docker Compose
• Install Git
• git clone git@github.com:journeymonitor/control.git
• git clone git@github.com:journeymonitor/monitor.git
• git clone git@github.com:journeymonitor/infra.git
• cd infra/docker
• docker-compose up

This gives you two running Docker containers, journeymonitor-control and journeymonitor-monitor.

• Within the control container, /opt/journeymonitor/control is a mirror of the root directory of the control git clone.
• Within the monitor container, /opt/journeymonitor/monitor is a mirror of the root directory of the monitor git clone.

If you are using Docker for Mac, you can access the web UI at http://localhost:8080.

If you are using Docker Machine, running echo "http://$(docker-machine ip):8080" will print the URI you need to open.

Setting up a production environment

This has been tested to work on an Ubuntu 14.04 64bit system.

• In puppet/hieradata/host, add a file named <fqdn of your host>.yaml.
• In this file, declare env: prod and role: all.
• Commit and push.
• On the target system, run sudo apt-get update && sudo apt-get install puppet.
• Then, clone this repository: git clone git@github.com:journeymonitor/infra.git.
• cd infra/puppet
• Run ./apply.sh.

This will result in the control and monitor applications being installed and set up after some minutes, will set up continuous delivery for these apps, and will serve the JourneyMonitor platform on port 80.

Setting up Continuous Delivery with TravisCI and SimpleCD

The following describes how we realize Continuous Delivery from all JourneyMonitor repositories to our production systems at http://journeymonitor.com.

• At https://github.com/settings/connections/174611, enable TravisCI access for the organization that holds the repositories
• At https://travis-ci.org/profile, hit the Sync button, and wait for your GitHub organization to become available
• Switch to the journeymonitor organization and switch on all repositories
• At https://github.com/settings/tokens, generate a new token with public_repo and repo_deployment rights
• Copy the newly generated token to the clipboard
• Install the travis gem via sudo gem install travis
• Go to the root folder of each repository clone (infra, control, monitor, and analyze), and run travis encrypt GITHUB_TOKEN=<token-from-clipboard>
• Add the resulting secure line to the .travis.yml file of the according repository, like so:

env:
  global:
  - secure: "i/g95ZV29lj...M4ZocNL+yo="

• For each repository, set up a .travis.yml file that builds and tests the application for this repository
• As the after_success step, define - make travisci-after-success
• As a consequence, each successful TravisCI run will result in a new GitHub release being generated
• This results in tags that are named travisci-build-${TRAVIS_BRANCH}-${TRAVIS_BUILD_NUMBER}, and for each repository, a SimpleCD cronjob is defined (see https://github.com/journeymonitor/infra/blob/puppet/modules/cronjobs/templates/etc/cron.d/journeymonitor-control.erb for an example) that watches these tags and triggers a deployment whenever a new tag appears

Guide to the repository

Noteworthy Puppet modules

puppet/modules/env-mgmt is the place where infrastructure-wide environment variables are managed (which are, e.g., used by the applications).

Modules starting with app- provide the system configuration that needs to be in place for the application in question to being able to operate - however, these modules do not install the application itself (this is achieved via the SimpleCD continuous delivery setup).

puppet/modules/cronjobs centralizes cronjob configuration - it uses hiera data from other namespaces (e.g. app-analyze::) in order to configure cronjobs for different apps.