DEVOPS WEEKLY ISSUE #157 - 5th January 2014
Lots of content this week as everyone seems to have spent at least some of the time over the holidays hacking on tools or writing blog posts. Anyone thinking that the new year is a great excuse to get back to writing regularly is right.
Sponsor
Devops Weekly is sponsored by Brightbox Cloud - launch cloud servers in multiple UK datacentres in seconds…
http://brightbox.com
News
Flapjack is basically a router for alerts and notifications from a monitoring system. It’s not trying to do everything but does one thing very well. This post explains where it fits and why you might want to give it a try.
http://fractio.nl/2014/01/03/the-how-and-why-of-flapjack/
A post all about the changing landscape of network hardware, as generic server hardware becomes much more capable of being used for switches and routers. This is interesting to me as it likely makes software defined networking easier, even with existing tools.
http://etherealmind.com/server-performance-network-agents-software-routers-and-networking/
Automation is not enough without adopting infrastructure as code practices, or so this post claims. A well reasoned argument that simply throwing automation, especially packaged products, at a problem will cause problems without a software mindset.
http://kief.com/infrastructure-as-code-versus-automation.html
The core language parser in Puppet has been getting some serious love for a while now on the road towards Puppet 4, and this blog already has several detailed posts about the new Type system.
http://puppet-on-the-edge.blogspot.co.uk/
A sort of ode to the role of a modern operations engineer. A great description of why the combination of typical operations activities and software development are fulfilling and attractive to people.
http://www.opsbs.com/2013/09/why-i-infracode/
A nice presentation about the changing role of testing in organisations that are shipping software very frequently.
http://www.slideshare.net/AlexSchwartz1/dev-opsberlin-ignitehelpwehavenomoretimefortestingslides
A description of the research techniques behind the recent Puppet Labs Devops survey, and some highlights from the findings of last years version. Definitely worth using when convincing people of the benefits of a devops approach.
http://continuousdelivery.com/2013/12/the-science-behind-the-2013-puppet-labs-devops-survey-of-practice/
I’ve been playing a bit with Serverspec for testing infrastructure components recently. This post, and accompanying repo, demonstrates testing Packer created images for AWS as part of a deployment pipeline.
http://www.morethanseven.net/2014/01/01/testing-packer-created-images-with-serverspec/
An argument that continuous delivery is now mainstream. I’m not sure silicon valley companies like Google, Facebook and Amazon are much like your typical enterprise, but it’s a good collection of links and talks about shipping rapidly at scale.
http://infiniteundo.com/post/71540519157/continuous-delivery-is-mainstream
Events
Early bird tickets are now on-sale for Craft in Budapest in April. Everything from development process, testing, continuous integration (I’m speaking about continuous integration for infrastructure), distributed systems and scaling to name a few topics.
http://craft-conf.com/2014/
Tools
A very nice tool that provides code coverage measurements for Chef Spec tests. The tooling around testing infrastructure code is becoming more and more mature every month it seems.
https://sethvargo.com/chef-recipe-code-coverage/
Another nice looking graphite dashboard. This one provides some structure to metrics stored in graphites and allows querying via GEQL, Graph Explorer Query Language.
http://vimeo.github.io/graph-explorer/
Fig is like a combination of vagrant and docker, you define the services which comprise your application in a YAML file and then type fig up to boot everything in docker containers.
http://orchardup.github.io/fig/