2 minute read

A relatively short issue this week as I’m writing this in the airport before flying out to Santa Clara for Velocity and Devopsdays. Which should mean lots of content for next issue to make up for it. Any readers going along to the above conferences do come and say hello.

Sponsor

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

News

A good post extolling the benefits of post-mortems for incidents, and in particular focusing on the need to make them blameless - and the knock on effect that has on an organisation’s culture.
http://www.paperplanes.de/2014/6/20/what-blameless-postmortem-taught-me.html

An excellent technical walkthrough of building a risking engine to detect fraud using machine learning techniques.
http://nerds.airbnb.com/architecting-machine-learning-system-risk/

And another post about using machine learning to process data sets, this site in pursuit of better recommendations. It feels like open source tools exist for the low level parts here, but not the higher level frameworks.
http://codeascraft.com/2014/06/18/conjecture-scalable-machine-learning-in-hadoop-with-scalding/

Not all of the excitement around docker comes from people who understand what docker is so I liked this post that aims to clear up a number of commonly seen misconceptions.
https://devopsu.com/blog/docker-misconceptions/

A blog post making that case that those that operate systems are users too. A good detailed example is talked through about why it’s important to have real users, in this case systems administrators, on the product team.
http://blog.jaywayco.co.uk/sysops-are-end-users-too/

A nice practical look at a few areas that many people would associate with devops;. collaboration, automation, continuous integration, continuous testing, continuous delivery and continuous monitoring.
http://blog.newrelic.com/2014/06/12/divingintodevops/

With the growth of monitoring and other times of time series data these requirements for a good time series database make for an interesting read.
http://www.xaprb.com/blog/2014/06/08/time-series-database-requirements/

Organising work is an interesting topic, and the last few years has seen lots of interest in Kanban for engineering and operations tasks. I enjoyed this next post mainly as it shows someones actual board. It would be interesting to compare lots of these.
http://matschaffer.com/2014/06/rethinking-kanban/

A long list of projects that should be of interest to any systems administrators out there. Lots you’ll probably be familiar with but I bet most people will find a few things they didn’t know about.
https://github.com/kahun/awesome-sysadmin

Tools

If you’ve ever looked into ways of securing traffic between internal services you’ll have found lots of options, most of them lots of work. Certified is an internal CA that’s designed to be easy to use
https://github.com/rcrowley/certified

Updated: