3 minute read

I talked at PupppetConf about how I think we’re still edging slowly towards infrastructure as code. A few of the links before point in that direction, with automated test suites for black box services, interactive ways of exploring infrastructure and tools for learning to code. More of the same please.


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


A good write-up from my old colleagues about how they approached writing acceptance tests for a Content Distribution Network. Black box testing like this feels useful for a bunch of different services or devices.

Devops owes lots to the discussion of agile that came before it, one of those discussions being around the importance of cross functional teams. This post stresses why having a real team is so important.

This next post talks about the importance to code review, and about making that easier with pre-checks like static analysis. The reason I like it is because it shows how simple this can be in trivial cases. Easy to forget when moving to describing infrastructure as code.

If you’ve managed a reasonable sized AWS infrastructure you’ll probably have spent at least some time looking for an instance based on incomplete data. Skew sounds an interesting solution, I’m particularly interested in seeing the IPython Notebook integration.

It’s been a week or since PuppetConf 2014 now so most of the slides are now available online. That’s nearly 100 presentations to keep you busy.

Quite a few nice points in this interview with RedHat CEO putting the business spin on devops. Interesting following the CEO/CIO discussions around something that is so community driven.

A follow up to a post I included a few weeks back, all about why leading people and management could be for you. Managers play a big part in how culture changes in most organisations and it’s often ignored by technology driven employees.



Akvo.org is a non-profit operating from Amsterdam. We work on increasing visibility for charities and their work, and provide survey and monitoring software in developing countries, especially in Africa and India. We’re searching for an infrastructure engineer ideally versed in Puppet, but in general what we want is someone who knows how to automate provisioning and build continuous delivery pipelines. Experience with Python or Java/Clojure is a plus. The main office is in Amsterdam, but the developers mostly work remotely; we’d prefer someone to be in one of our hubs (Amsterdam, Stockholm or Helsinki) but remote ±2 hours of CEST can work too.

[email protected]


Join us in Seattle, WA, November 9-14, 2014, for LISA14.

The LISA conference has long served as the annual vendor-neutral meeting place for the wider system administration community. Recognizing the overlap and differences between traditional and modern IT operations and engineering, the highly-curated 6-day program offers training, workshops, invited talks, panels, paper presentations, and networking opportunities around 5 key topics: Systems Engineering, Security, Culture, DevOps, and Monitoring/Metrics. Don’t miss the chance to be a part of this unique career-building journey.

Register by October 20 and save: https://www.usenix.org/lisa14/dow


Chef Guard looks like a useful tool, specifically it protects your Chef server from untested and uncommitted cookbooks. Several useful features, including emailing of configuration change diffs and validation of changes.

Learning new tools and languages is always hard so anything that makes it easier it a good idea. Chef Koans takes inspiration from other similar projects and applies it to Chef, a good starting point if you’ve done similar learning exercises before.

Zonify is a set of command line tools for managing Route53 DNS for an AWS infrastructure. It intelligently uses tags and other metadata to automatically create the associated DNS records.