DEVOPS WEEKLY ISSUE #154 - 15th December 2013

You would think people would be starting to wind down for the holidays but the last few weeks have seen loads of interesting content. Lots of infrastructure testing tools and posts, case studies and examples of deployment automation, what network engineers should be learning next year and a survey you should all fill in about security in the development process.

Sponsor

Devops Weekly is sponsored by Brightbox Cloud - launch cloud servers in multiple UK datacentres in seconds…
http://brightbox.com

News

Here’s a good post about why writing tests against your infrastructure is a good idea, along with a pretty interesting approach to testing load balancers using the honeyd honeypot network daemon.
http://mauveweb.co.uk/posts/2013/12/load-balancer-testing-with-honeyd.html

A survey for developers, operations engineers and security folk all about who is responsible for different aspects of security within your development team. Please fill this in as the results should be interesting when they’re available.
https://www.surveymonkey.com/s/Developers_and_AppSec

Really interesting look at what adoption of Devops approaches in a typical enterprise organisation might look like. Makes lots of good points about why this is different, and makes some comparisons with agile adoption.
http://www.slideshare.net/ewolff/continuous-delivery-and-devops-in-the-enterprise

Writing tests makes your code more maintainable, but testing command line tools can be tricky and not everyone bothers. This post shows a neat approach for making testing the effects of command line tools easier, even when they make both local and remote changes.
http://holmwood.id.au/~lindsay/2013/12/06/cli-testing-with-rspec-and-cucumber-less-aruba/

A write-up of different approaches to service discovery, and an introduction to the new SmartStack framework. A good description of the ideal system, along with why approaches based on DNS, central load balancing, and in-app registration don’t always work.
http://nerds.airbnb.com/smartstack-service-discovery-cloud/
https://github.com/airbnb/synapse
https://github.com/airbnb/nerve

The changes in networking as more things are virtualised are interesting, and this presentation does into the skills needed in the future. So from CLIs and closed source to APIs, open source and software defined networking.
http://www.slideshare.net/mestery/next-gennetworkengineerskills

Blog post, with lots of screenshots, of how Yammer deploy software. Lots of nice ideas and a focus on self service deployments.
http://eng.yammer.com/deployments-at-yammer/

Similar to the Yammer post, here’s the story of moving to automated deployments at The Trainline. Some tips for using Thoughtworks Studio’s Go pipelines but also good points about business value and culture.
http://engineering.thetrainline.com/2013/12/02/moving-to-multiple-deployments-per-week-at-thetrainline-com/

An interesting discussion about the common params.pp approach for managing data in Puppet modules, along with a proposal and working implementation of an alternative approach using hiera data.
http://www.devco.net/archives/2013/12/08/better-puppet-modules-using-hiera-data.php
http://www.devco.net/archives/2013/12/09/the-problem-with-params-pp.php

Jobs

RethinkDB is an open-source distributed database built to store JSON documents at scale. Our software is a complex system that allows execution of custom code on arbitrary datasets and your job, as a reliability engineer, is to ensure RethinkDB clusters are always fast and stable, regardless of what the user throws at them. Your primary responsibilities will finding, fixing, and preventing performance bottlenecks in RethinkDB and building testing automation infrastructure.
http://rethinkdb.com/jobs/reliability/

Events

Scale Summit is an unconference event, bringing together professionals from the operations and software development communities who have a particular interest in scalable, high performance systems. It’s going to be in London early next year and you can sign up now to find out when more information is available.
http://www.scalesummit.org/

Tools

Bosh, the release engineering toolkit from CloudFoundry, can be used for all sorts of management tasks, but it’s not the easiest tool to learn. This project provides a nice example of standing up a Chef managed stack using Bosh which aims to be easier to follow that standing up an entire CloudFoundry instance.
https://github.com/hatofmonkeys/bosh-chef

Chef browser is a nice web interface for information in your Chef server. If you use knife search and knife show a lot this might be of interest.
http://3ofcoins.net/2013/12/11/we-released-chef-browser-today/
https://github.com/3ofcoins/chef-browser

Multi-Mechanize is an open source framework for performance and load testing. It runs concurrent Python scripts to generate load against a remote site or service. The nice part is it’s very much based around a unit testing assertion approach.
http://testutils.org/multi-mechanize/

If you’re wanting to test outgoing port firewall rules you need something to make a request of that is listening on that port. Enter portquiz. Combining this with Serverspec would be great.
http://portquiz.net/