amazon web services - cloudinit and route53 bootstrap
:::::::: ::: ::::::::: :::::::::
:+: :+: :+: :+: :+: :+: :+: :+:
+:+ +:+ +:+ +:+ +:+ +:+ +:+
+#+ +#++:++#++: +#++:++#: +#++:++#+
+#+ +#+ +#+ +#+ +#+ +#+ +#+
#+# #+# #+# #+# #+# #+# #+# #+#
######## ### ### ### ### #########
- cloudinit and route53 bootstrap -
Install
rubygems
gem install aws-carb
Dependencies
If you've compiled your own ruby (rvm/rbenv, etc) then you will need to make sure ruby has support for various libraries.
For debian based systems, do the following before compiling ruby:
sudo apt-get isntall libreadline-dev zlib1g-dev libssl-dev
Configuration
At minimum, Carb needs to know your aws-ec2 credentials. The simplest way to allow this is to edit ~/.carb/config/config.yaml. See config/config.yaml.example for an example config.
Example usage
carb can be used to create ec2 instances
aws create
Temporary Quickstart Docs..
Leaving this stuff here until proper docs have been written:
# works on ruby > 1.9
gem install aws-carb
# install config and template directories
mkdir -p ~/.carb/{config,template}
sudo updatedb
cp `locate gems/aws-carb-0.0.3/examples/config/config.yaml.example` ~/.carb/config/config.yaml
cp `locate gems/aws-carb-0.0.3/examples/template/basic.cloud-config.erb` ~/.carb/template/
# edit config.yaml - remember to specify an ssh key because without one your ec2 instance will be inaccessible! (copy your ssh key from id_rsa.pub as a string - not a file path)
vim ~/.carb/config/config.yaml
# examples:
# create a basic instance like this (or put 'image_id' into 'ec2' section of config.yaml):
carb create --image-id <ami id>
# create an instance but bootstrap it with the user-data cloudinit template and also get some route53 goodness going on because hostname is set
# run with verbose to see more interesting info
# Note: parameters specified on the command line override variables set in the config file
carb -v create --user-data-template ~/.carb/template/basic.cloud-config.erb --common-variables "{ 'hostname' => 'asdasdasasdasdsa' }"
# list all the ec2 options:
carb help create
Understanding block-device-mappings
Description
When specifying block-device mappings, each mapping must be a valid hash and must be contained in an array.
Once the machine has booted the block device(s) will be available through their virtual block device, e.g:
# /dev/xv<disk><partition>
/dev/sdc1 = /dev/xvc1
/dev/sdf0 = /dev/xvf0
...
http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/block-device-mapping-concepts.html
Examples
A typical block device mapping:
[{ 'device_name' => '/dev/sdf1', 'ebs' => { 'volume_size' => 500 } }]
Multiple block device mappings:
[
{ 'device_name' => '/dev/sdf1', 'ebs' => { 'volume_size' => 500 } },
{ 'device_name' => '/dev/sdf2', 'ebs' => { 'volume_size' => 500 } }
]
Typical command line usage:
carb -c ~/.carb/config/config.yaml -v create --user-data-template ~/.carb/template/basic.cloud-config.erb --block-device-mappings "[{ 'device_name' => '/dev/sdf1', 'ebs' => { 'volume_size' => 500 } }]"