3 minute read

I’ve been on holiday this week so it’s a relatively short issue, but lots of interesting bits and pieces to read. Next Sunday is another attempt to write the newsletter from an airport, so if you don’t get it around the usual time then it should arrive sometime Monday when I get to Portland, Oregon.


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 very indepth look at kernel debugging with ftrace. Ftrace is likely already built-in and enabled if your using Linux but not something I’ve seen used by many people.

A good case for people to stop sub-labelling devops, for instance as enterprise devops or similar. Ultimately we’re better learning from as wide a group of experiences as possible it seems to me.

Microservices are definitely hip, but not enough people are talking about the challenges in my experience. This post does a good job of pointing out complexity still exists, you’re just moving it elsewhere. That elsewhere is often why stronger operations skills are in such demand.

A nice collection of presentations from the Papers we Love meetup group. Lots of talks by different people in different cities, but a good start if you’re starting to look at academic literature.

Having a tool to help guide your programming style is incredibly useful, both when learning a language and when aiming for consistency between different projects. Puppet Lint is just that tool for the Puppet language and it’s just hit a 1.0.0 release, with the addition of a pluggable rules and the ability to automatically fix some issues.

Another deep-dive into AWS performance, this time looking at Route 53 DNS and in particular the new geolocation routing policy vs a latency based policy.


Say Media (digital publishing and technology company based in SF) is looking for a seasoned Operations Engineer to help us manage, maintain, and deploy key infrastructure. As a member of the Ops team, you’ll be responsible for maintaining, improving, and optimizing a global environment. Please check out the job description at:

Digitas LBi is a digital rebranding agency in the Truman Brewery, Brick Lane. We’ve recently merged with Digitas and the Publicis group and our technology function is undergoing an exciting stage of growth and change. We’re looking for a DevOps evangelist to join the brand new DevOps function to set up processes and revolutionize the way we work across Managed services and Technology. You’ll be a pro with Puppet, Chef, vagrant, AWS and live and breathe TDD.
https://careers-digitaslbi.icims.com/jobs/2195/senior-devops-engineer/job Contact Kirsty Nelson for more details: [email protected]


The London Devops meetup is back, with a few short talks ands lots of time for conversations. The first event will be on September 9th at 7pm at the Just Eat offices with the second event already in the works for October.


Mentioned in the ftrace blog post above this set of scripts is a great resource if you’re doing any low-level performance tuning on Linux.

Vessel is another tool for automating the setup and use of docker based development environments. It integrates Vagrant, CoreOS, Consul, docker and Atom into a slick opinionated package. The user interface looks particularly impressive from the screenshots.

Ganger is a tool for running short-lived network services inside Docker containers, and then proxying client connections to them. It’s designed mainly for data processing batch jobs that require a fresh environment each time, but could have other interesting uses.

Chef-runner is a tool for executing chef recipes on remote machines, over SSH. It features vagrant integration and is built to be much faster than the in-built provisioning for this use-case.