2 minute read

It’s the second Monitorama conference this coming week so hopefully lots of monitoring related content next weekend. If the last one in Boston was anything to go it should be a pretty special event. If you’re going to be in Berlin let me know on twitter @garethr, I’m arriving Tuesday.


Devops Weekly is sponsored by Brightbox Cloud - launch cloud servers in multiple UK datacentres in seconds…


A presentation from Surge all about engineering management. Loads of good content in here about organisational structures which tend to put software developer off your company, as well as discussion of incentive and job titles.

A great story of optimising a web stack. Stepping through an existing architecture, load testing to find bottlenecks and then the changes that worked and didn’t.

Good video of a talk all about notable devops culture traits. Good examples from specific companies, including Netflix and IBM, as well as a quick introduction to Deming.

Blog post about integrating a bunch of monitoring tools together, in this case statsd, graphite, nagios and sentry. A good tale of increasing the visibility of a system, making finding and fixing issues quicker and easier.

Managing secrets when using configuration management tools is important, and this detailed walk through of using chef-vault is a great starting point for any Chef users. Lots of code examples and demonstrations of using the knife plugin.

A quick survey of the state of package management tools and how people are using them. I’ll hopefully post a link to the results too once they are published.

If you have an interest in distributed systems then you’ve probably found yourself reading an academic paper or two. This curated list looks like an excellent starting point for a deep dive.

An interesting approach to internal service communications using AMQP, and specifically RabbitMQ. The post describes a Ruby implementation and shows some examples, but the conventions could be adopted by any language or framework.

A useful diagram highlighting that different parts of the stack, from hardware and OS up to application code and configuration, change at different rates. I’d love to see actual real world data on this to back up the hypothesis.

Nice writeup of the advantages of adding testing to configuration management code, in this case focusing on the evolution of the Riak chef cookbook and the testing tools used.


London based DevOps consultancy The Scale Factory are looking to expand their team. If you’d like to work on challenging infrastructure problems across a range of different businesses, you can find more details at http://www.scalefactory.com/team/careers or you can find @jtopper on Twitter or ##infra-talk”


Ventriloquist is a new tool which combines vagrant and docker to make building development environments with dependencies easier. It’s implemented as a vagrant provisioner which can make use of a bunch of pre-build docker images t provide common services.

Batou looks like an interesting tool for service deployment. It aims to mix the simplicity of tools like Fabric with a more model driven idempotent system like Puppet or Chef.

Collectd3 is a tool which takes information from collectd and visualises it using the D3 library, providing some nice graphics as well as a bird-eye view of multy-server system.