Stackbuilder Config Example

Stackbuilder Config is a DSL that is used by stackbuilder to configure allocate, provision 'Stacks' of Virtual Machines This repository provides example configuration for this DSL.

Terminology

Virtual Machine:

• A single machine of a given type: e.g a single apache server
• Can be individually configured but are generally identical to their siblings

Machine Set:

• A simple container consisting of 1 or more Virtual Machine's
• Machine Sets can be extended take on a custom behaviour: eg. MySQL Cluster, RabbitMQ cluster, App Service, Mail Service, Load Balancer Service.
• The container has all the meta information and attributes that apply to all Virtual Machines in the container.
• Eg. A machine set extended by App Service could contain the name of the application, ports, firewall rules and the number of instances(VM's).

Stack:

• Container of 1 or more Machine Set(s).
• Can be bound to 0 or more environemnts
• Can be configured differently but are generally identical across all environments (consistency)

Site:

• A specific geographic location.
• Typically referenced by an abbreviated name.
• Eg. 'lon' (London) or 'ny' (New York)

#####Environments:

• A container of 'Stacks' in 1 or 2 sites
• Each environment has a primary and an optional secondary site.
• The Machine Set is responsible for determining which site Virtual Machines should reside.
• Environments can also contain other environments (more advanced topic).
• The environment can also include meta information such as networking routes that apply to all 'Stacks' in that environment.

Computenode:

• This is a dom0 (KVM host) that has been configured to host stack built Virtual Machines
• Dependencies:
  • provisioning-tools
  • Dynamic DNS
  • MCollective

Overview

---

Usage information

###Virtual Machines

####Show Virtual Machine spec for Virtual Machine staging-helloapp-001

$ env='staging' rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:to_specs
      - !ruby/sym networks:
      - !ruby/sym mgmt
      - !ruby/sym prod
    !ruby/sym ram: "2097152"
    !ruby/sym fabric: earth
    !ruby/sym domain: earth.net.local
    !ruby/sym vcpus: "1"
    !ruby/sym qualified_hostnames:
      !ruby/sym mgmt: staging-helloapp-001.mgmt.earth.net.local
      !ruby/sym prod: staging-helloapp-001.earth.net.local
    !ruby/sym hostname: staging-helloapp-001
    !ruby/sym availability_group: staging-helloapp
    !ruby/sym storage:
      !ruby/sym /:
        !ruby/sym type: os
        !ruby/sym prepare:
          !ruby/sym method: image
          !ruby/sym options:
            !ruby/sym path: /var/local/images/gold/generic.img
        !ruby/sym size: "5G"

####Show ENC data (puppet code) for a Virtual Machine staging-helloapp-001

env='staging' rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:to_enc
---
  "role::http_app":
    port: "8000"
    dependencies:
      db.hello.hostname: staging-hellodb-001.earth.net.local
      db.hello.database: hello
      db.hello.username: JavaHttpRef
      db.hello.password_hiera_key: enc/staging/JavaHttpRef/mysql_password
    environment: staging
    vip_fqdn: staging-helloapp-vip.earth.net.local
    group: blue
    dependant_instances:
      - staging-helloproxy-001.earth.net.local
      - staging-helloproxy-002.earth.net.local
      - staging-lb-001.earth.net.local
      - staging-lb-002.earth.net.local
    application: JavaHttpRef

####Show all options available for the Virtual Machine staging-helloapp-001

$ env='staging' rake -T staging-helloapp-001.mgmt.earth.st.net.local
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:allocate            # allocate these machines to hosts (but don't actually launch them - this is a dry run)
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:allocate_vips       # allocate IPs for these virtual services
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:audit_hosts         # new hosts model auditing
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:clean               # clean away all traces of these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:disable_notify      # disable notify for these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:enable_notify       # enable notify for these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:free_ip_allocation  # frees up ip and vip allocation of these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:free_ips            # free IPs
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:free_vips           # free IPs for these virtual services
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:launch              # launch these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:mping               # perform an MCollective ping against these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:orc:resolve         # deploys the up2date version of the artifact according to the cmdb using orc
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:provision           # perform all steps required to create and configure the machine(s)
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:puppet:clean        # Remove signed certs from puppetmaster
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:puppet:poll_sign    # sign outstanding Puppet certificate signing requests for these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:puppet:run          # run Puppet on these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:puppet:sign         # sign outstanding Puppet certificate signing requests for these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:puppet:wait         # wait for puppet to complete its run on these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:reprovision         # perform a clean followed by a provision
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:resolve             # resolve the IP numbers of these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:test                # carry out all appropriate tests on these machines
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:to_enc              # outputs the specs for these machines, in the format to feed to the provisioning tools
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:to_specs            # outputs the specs for these machines, in the format to feed to the provisioning tools
rake sbx:staging-helloapp-001.mgmt.earth.st.net.local:to_vip_spec         # outputs the specs for these machines, in the format to feed to the provisioning tools

Stacks

####Show all stacks available with the name 'helloworld'

$ env='staging' rake -T helloworld
rake sbx:helloworld:allocate            # allocate these machines to hosts (but don't actually launch them - this is a dry run)
rake sbx:helloworld:allocate_vips       # allocate IPs for these virtual services
rake sbx:helloworld:audit_hosts         # new hosts model auditing
rake sbx:helloworld:clean               # clean away all traces of these machines
rake sbx:helloworld:disable_notify      # disable notify for these machines
rake sbx:helloworld:enable_notify       # enable notify for these machines
rake sbx:helloworld:free_ip_allocation  # frees up ip and vip allocation of these machines
rake sbx:helloworld:free_ips            # free IPs
rake sbx:helloworld:free_vips           # free IPs for these virtual services
rake sbx:helloworld:launch              # launch these machines
rake sbx:helloworld:mping               # perform an MCollective ping against these machines
rake sbx:helloworld:orc:resolve         # deploys the up2date version of the artifact according to the cmdb using orc
rake sbx:helloworld:provision           # perform all steps required to create and configure the machine(s)
rake sbx:helloworld:puppet:clean        # Remove signed certs from puppetmaster
rake sbx:helloworld:puppet:poll_sign    # sign outstanding Puppet certificate signing requests for these machines
rake sbx:helloworld:puppet:run          # run Puppet on these machines
rake sbx:helloworld:puppet:sign         # sign outstanding Puppet certificate signing requests for these machines
rake sbx:helloworld:puppet:wait         # wait for puppet to complete its run on these machines
rake sbx:helloworld:reprovision         # perform a clean followed by a provision
rake sbx:helloworld:resolve             # resolve the IP numbers of these machines
rake sbx:helloworld:test                # carry out all appropriate tests on these machines
rake sbx:helloworld:to_specs            # outputs the specs for these machines, in the format to feed to the provisioning tools
rake sbx:helloworld:to_vip_spec         # outputs the specs for these machines, in the format to feed to the provisioning tools