3 minute read

Definitely the first issue of devops weekly where the first three news items are Windows related. At the recent Build conference it looks like Microsoft is making a genuine attempt to catch up with the last few years of Infrastructure as code - and in particular integrating with existing popular tools like Chef, Puppet, Vagrant and Chocolatey. This can only be a good thing in my view.


Devops Weekly is sponsored by Brightbox Cloud - serious UK-based cloud infrastructure from only 1.5p per hour (£10.95/month)

Start your £20 free trial now: http://brightbox.com/devopsweekly


Microsoft have released Windows Management Framework V5 preview, with the oddly named OneGet. This is a simple package manager for PowerShell which can install packages from the community Chocolatey repository. Intriguingly “Support of additional repositories will come in subsequent versions”.

Microsoft have also released an official Windows vagrant box, of Windows Server 2012 R2 Datacenter. This appears to be a trial, with the download available for 90 days and the box working for 180 days.

More news for Windows users. At the recent Build conference both Puppetlabs and Chef announced much improved Windows and Azure support. Combined with the above Vagrant box it feels like Windows might be starting to catch up on the automation front.

My talk at Puppetcamp London this week, all about test driven development with Puppet. It’s mainly aimed at being an introduction to software testing, test driven development and then to the tools available for Puppet - including the new Beaker acceptance testing framework.

Another of the talks from Puppetcamp London, this one all about building, testing and using AMI images on EC2. Lots of practical tips from using Jenkins pipelines to ‘immutable’ servers.

The service management framework ITIL has many fans, and this is a really nice writeup of how it may change based on devops experiences

A post all about why Vagrant is such an important tool when adopting a devops mentality. By having a tool that everyone uses you can cut across silos and ultimately build empathy.

A good set of recommendations around branching strategies, including feature branching in code rather than using your source control system. Lots of folks still resist doing everything in master.

A good post comparing a deployment specific tool (in this case Octopus) with a configuration management tool for deployment. Some good observations about similarities and differences. The comments also add to the debate about whether one or two tools is best.

Nginx is one of my favourite bits of software, but historically you’ve had to spend at least some time learning the basics from a variety of different source. The new official admin guide is therefore pretty handy for those that haven’t used Nginx yet.


deviantART is looking for a DevOps engineer, remote or local to Los Angeles, to help run our infrastructure that serves over 2 billion pageviews per month. We use puppet, graphite, python, and phabricator to automate, monitor, virtualize, and scale everything. If you would like to join our small but talented team, apply at:

Eventbrite is the world’s fastest-growing online ticketing and registration platform. We’re looking for a Sr. Systems Engineer to join our small team using Puppet, Django and other automation tools to manage and scale our site as our company continues to grow. We’re constantly experimenting with new ways to monitor and maximize site performance. Sound up your alley?

Adobe’s Digital Publishing Services team is looking for a top notch Sr. Site Reliability engineer to work within our development team to create and operate our mission critical systems. Our rich multi-platform solution is widely used by the world’s largest publishers and is seeing strong growth within enterprises.


Promise 2014 is the CFEngine user conference, being held in San Francisco from September 17th to 19th. Early bird tickets are available now and a call for proposals for talks is also open.


Sysdig is a new Linux system troubleshooting tool, described as a combination of strace, tcpdump and lsof. The examples are pretty impressive and you can imagine lots of tools you could build on top of this.

Mentioned at Puppetcamp London, Gonzo is a dashboard for Puppet focused on assessing the impact of changes made by Puppet. A clever way of bridging a traditional operations model with infrastructure as code.