To start using Serverless and access the AWS API you need to set the AWS API Access Key & Secret.
As a quick setup to get started you can export them as environment variables so they will be accessible to Serverless and the AWS SDK in your shell:
export AWS_ACCESS_KEY_ID=<key>
export AWS_SECRET_ACCESS_KEY=<secret>
| cd | |
| rm -fr go | |
| wget http://go.googlecode.com/files/go1.2.linux-amd64.tar.gz | |
| tar xfz go1.2.linux-amd64.tar.gz | |
| cd clone | |
| go get github.com/kr/godep |
| module exports = class SomePlugin extends ServerlessPlugin{ | |
| //Contructor should be defined in ServerlessPlugin so you don't have to recreate it here | |
| constructor(pluginConfig, serverlessConfig, serverlessUtils) { | |
| this.pluginConfig = pluginConfig; | |
| this.serverlessConfig = serverlessConfig; | |
| this.serverlessUtils = serverlessUtils; | |
| } | |
| beforeDeployResources(params) { |
| # Service Interruption Debrief | |
| ## Brief Executive Summary | |
| EXECUTIVE SUMMARY HERE | |
| ***On Call during outage***: | |
| ***Time outage started***: |
| #!/usr/bin/env ruby | |
| require "open3" | |
| git_branch = ENV['CI_BRANCH'] || `git rev-parse --abbrev-ref HEAD` | |
| @app_name = "csqa-#{git_branch[0..24].gsub(/\W/, '-')}" | |
| def exec command | |
| Open3.popen2e(command){ |stdin, stdout_err, wait_thr| | |
| while line = stdout_err.gets |
| *) Ask them if they ever spend time manually testing features that have been out for a while | |
| *) Ask them if they ever shipped new features that broke old features | |
| *) Ask them if they are happy with this? | |
| *) Ask them if they think it makes them and their company successful or if they think they need to change | |
| You can take a couple of points from the business case blogpost about why to do CI here: | |
| http://blog.codeship.io/2013/04/11/a-business-case-for-continuous-integration.html | |
| Testing is about making sure your users can walk through the most important parts of your system and use your product. When testing you should be thinking about your users first and how they interact with your system, not how you can get 100% unit test coverage (which is a nonsense metric) |
| If you only want to send a hook for successful builds at the end of your test commands add something like: | |
| wget YOUR_URL | |
| and it will send the hook whenever your build passes. You can run any command there and aren't limited to wget or anything at all. | |
| 2) Send a hook for a failed command and fail the build: | |
| SOME_COMMAND || bash -c "ANY_COMMAND_TO_SEND_HOOK; exit 1" |
| phpunit -d xdebug.max_nesting_level=250 | |
| # or | |
| echo "xdebug.max_nesting_level=250" >> ~/.phpenv/versions/$php_version/etc/php.ini |
| mkdir .casperjs | |
| cd .casperjs | |
| wget --retry-connrefused -q -O - https://codeload.github.com/n1k0/casperjs/legacy.zip/1.1-beta1 > casperjs.zip | |
| unzip casperjs.zip | |
| mv n1k0-casperjs*/* ./ | |
| cd .. | |
| # Add .casperjs to the beginning of the path so it takes 1.1-beta1 | |
| export PATH=/home/rof/clone/.casperjs/bin:$PATH |
| # Creating the EC2 Instance | |
| instance = ec2.instances.create( | |
| :image_id => "ami-7739b41e", | |
| :instance_type => 'cc2.8xlarge', | |
| :key_name => "CloudshipCreateServerKey", | |
| :instance_initiated_shutdown_behavior => 'terminate', | |
| :block_device_mappings => { | |
| "/dev/sda1" => { | |
| :volume_size => 50, |