lusheeta

This set of scripts helps you to manage separate clusters in your cloud. The tool lets you create and cleanup a cluster of your choice (see #Configuration).

---

Usage

$ ./main.py -h
usage: main.py [-h] -a {create,cleanup,prepare_ansible,run_ansible}
               [-c CONFIG] [-v]
               project

Cloud CLI tool

positional arguments:
  project               the name of the project

optional arguments:
  -h, --help            show this help message and exit
  -a | --action {create,cleanup,prepare_ansible,run_ansible}
                        the action to do
  -c | --config CONFIG
                        path to the configuration file
  -v, --verbose         set verbosity mode

The following examples will show how to setup a cluster (after configuring all necessary settings in the config file):

    # create cluster with an external config.yml + show debug messages
    $ ./main.py --action create --config /path/to/config.yml -vvv myproject
    
    # after cluster has been created, generate necessary ansible files + show only 'info' level logging
    $ ./main.py --action prepare_ansible --config /path/to/config.yml -vv myproject
    
    # run ansible playbook to setup software infrastructure
    $ ./main.py -a run_ansible -c /path/to/config.yml myproject
    
    # cleanup cluster from the cloud
    $ ./main.py -a cleanup -vvv myproject

---

Configuration

You can find the configuration files in the ./config dir.

default.yml - main config

This configuration file is the main configuration for the CLI tool. You can set the following parameters:

• project

  the name of the project (cluster) you want to manage.

• projects_dir - the path to the default projects directory

  • default: ./projects

• platform - the cloud platform of your choice

  • default: openstack
  • currently supported platforms: openstack

• platform_settings_file optional - the path to settings_file for the platform. It's the same as settings_file in supported_platforms.yml but if defined, has a higher priority (easier to define openstack credentials outside the source)

• network dict - the network settings for your cluster (when creating)

  • cidr - the CIDR for your cluster's network.
    • default: auto
    • possible values: auto or XXX.XXX.XXX.0/24
  • cidr_template - the CIDR template. Currently the script will substitute the X character from the template.
    • default: 10.X.100.0/24
    • Required when cidr == auto, otherwise optional
  • ext_net_name - the name of the gateway for your network to connect to the external network
    • default: ext-net

• vm_management dict - vm management settings for your cluster (when creating)

  • default_image_name - the default name of the image to use to spin up a vm. If you don't specify the image_name property for a host, this value will be used.
    • default: Ubuntu 14.04.2_20150505
  • default_vm_flavor - the name of the flavor to spin up a vm. If you don't specify the vm_flavor property for a host, this value will be used.
    • default: m1.medium
  • hosts_startup_timeout - the amount of time (in seconds) to wait to spin up all the vms and then keep executing the rest of the script
    • default: 600
  • terminate_vm_poll - the amount of time (in seconds) to wait between polls when terminating vms
    • default: 5

• hosts array - the section to setup host configs. Each array item is a host dict, which is a configuration for one host

  • name required - the name of the host

  • vm_flavor optional - the name of the flavor to spin up a vm. This property overrides vm_management.default_vm_flavor

  • count optional - the number of vms to spin up with these properties. If this field is missing, default value of 1 will be used.

  • image_name optional - the name of the image to spin up a vm. This property overrides vm_management.default_image_name

  • cloud_vars optional array - implementation specific special variables. Each item in the array must be a dict that contains the following parameter:

    • index all | <number> - the index to which host to apply the current special var

      • possible values can be all or the index <number> of the host
      • see config/default.yml as example

      Implemented options so far:

      • assignPublicIP boolean - when true, a public IP will be assigned to the index-th host

  • ansible_settings array - the section to setup the ansible settings for the host. Each item in the array is a dict with the following parameters:

    • ansible_group - which ansible group this host will belong to

    • group_vars array - an array of dict items to describe group-level variables. Each item must contain an index key and the desired parameter key with its value

      • index - which item in the list should have the parameter key
      • parameter_key: parameter_value - the parameter key-value to add in the inventory file for the current host

    • item_vars array - an array of key-value pairs that every item should contain in the group

• ansible dict - ansible settings to setup the software infrastructure

  • ansible_dir - the path to directory where your ansible project files reside
  • playbook - relative path to the playbook to run your setup
  • templates_path - path to folder that contains template files
  • inventory_template optional - a jinja2 template file for your inventory to use
  • ssh_config_template optional - a jinja2 template file for the ssh.config file
  • ansible_cfg_template optional - a jinja2 template file for the ansible.cfg file
  • ansible_bin_path required - the folder that holds ansible, ansible-playbook, etc

---

** For OpenStack networking the official OpenStack SDK was used as libcloud doesn't support the required functionalities (i.e. creating/removing subnets, networks, routers, etc).

UPDATE libcloud dropped completely.

---

Extension

./config/supported_platforms.yml contains the implementations of different platforms (so far only openstack is supported).