3 minute read

One blog post and a few tools this week confirming something I’ve been seeing too; traditional networking is definitely becoming more interesting to the devops community.


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


Lots of short hands-on screencasts aiming to teach lots of systems administration basics. Everything from man pages, command line tools, LXC containers and running postmortems.

A good reflections on Devopsdays Silicon Valley post, focusing on a observation I’d made as well, an increase in interest in all things networking at Devops events.

A nice devops in an enterprise presentation with a few nice ideas, especially for communicating intent to colleagues from other areas of the organisation. I particularly liked the pace of change visualisation and discussion.

A nice reminder that any topic of interest, and especially devops, is built on top of previous endeavours. Would devops exist without ITIL or the work of Deming? Learn from past work without being constrained by it.

A good call for developers to involve operations folks in the early stages of a project. My experience here is that devops is introduced from different places in different orgs, so the problem isn’t universal - but it’s still a good reminder.

A nicely documented example of building up a Jenkins continuous integration setup using ansible, and using docker to isolate test runs. Lots of code examples.

A pretty detailed post about when and whether to use software as a service over something a little more self-hosted, in particular looking at version control systems in the light of the recent Code Spaces episode.

A nice short deck on the importance of investing in your Puppet development workflow, going through a number of tools and processes and different approaches. Some good ideas for useful tools here too.


In Springer, one of the world’s leading scientific publishers, we are looking for Systems/Platform Engineers (and other roles) for London and Berlin. We are building the foundations to support our growing microservices landscape using IaaS, PaaS and other cool stuff. Try out our assessment! http://bit.ly/Tu5ft2

Uber is looking for a senior DevOps engineer to join the Realtime Platform team in San Francisco. The Realtime team is a small group of engineers who build, scale, and support the mission critical infrastructure running uber’s entire logistics platform. Ideal experience includes Python or Node.js, configuration management (like Puppet or Salt), alerting frameworks (like Nagios), NoSQL datastores (like Redis), haproxy, distributed task queues (like Celery).


The final program for FlowCon went up this week, including talks from Marty Cagan, Mary Poppendieck and Don Reinertsen, as well as talks on Google’s annual disaster recovery exercises, microservices and release engineering. Devops Weekly readers can get $50 off with the code DEVOPS50 too.


Haka is described as an open source security oriented language which allows to apply security policies on (live) captured traffic. It’s based on Lua and looks to be a great way of writing certain kinds of network tools.

Another interesting networking stack, again using Lua (LuaJit in this case). Snabb Switch is designed for high throughput, low latency network sytems suitable for use by ISPs. It speaks natively to ethernet hardware, hypervisors, and the linux kernel. Interesting times for networking.

An interesting approach to server authentication, swapping the AuthorizedKeysCommand for a custom application which checks keys published to GitHub rather than a local public key. The examples directory makes understanding how it works nice and easy.