DEVOPS WEEKLY ISSUE #156 - 29th December 2013
A short issue this week I’m afraid, but still a few interesting posts and tools to investigate in-between preparing for new year. If anyone writes up their thoughts about technology, systems administration or devops in 2013 do send me the links.
Sponsor
Devops Weekly is sponsored by Brightbox Cloud - launch cloud servers in multiple UK datacentres in seconds…
http://brightbox.com
News
Lots of interest and experiments in different ways of managing infrastructure at the moment. This example uses Docker and Mesos and provides a nice diagram of how everything links together.
http://tech.riywo.com/blog/2013/12/20/fluentd-on-mesos-plus-docker-plus-marathon/
Whether you’re using docker on top or just using LXC it’s worth understanding how the networking works. This blog post goes into lots of details, with code examples and tests taking you through most of the options.
http://containerops.org/2013/11/19/lxc-networking/
A pretty detailed blog post walking through implementing a Puppet provider, mainly focused on explaining what all the moving parts are in cases where the official documentation isn’t great.
http://garylarizza.com/blog/2013/12/15/seriously-what-is-this-provider-doing/
When discussing databases you often hear people talking about consistency. It’s both the C in CAP and the C in ACID after all. This post points out though that consistency in these two cases is actually a different thing. Worth considering next time you’re arguing about databases.
http://thinkrelevance.com/blog/2013/12/23/beware-inconsistent-definitions-of-consistency
Good Chef case study with a few interesting points. Talks about the advantages of adopting a single language (Ruby) for all operations tasks and a bunch of other team behaviours that helped.
http://www.slideshare.net/ulfmansson/chefrecordedfuture-29063408
Jobs
The UK Government Digital Service is looking for web operations engineers. With a strong systems background, you will ideally have a passion for developing the right tooling to create and evolve a resilient and secure platform on top of Infrastructure as a Service and enthusiasm for working in a cross-disciplinary devops environment. You will have excellent troubleshooting and research skills with a positive and proactive approach to customer service and getting things done.
https://gdstechnology.blog.gov.uk/2013/12/27/jobs-web-operations-engineer/
Tools
I’ve been experimenting with Serf recently but wanted a simple, testable way of writing easily distributed event handlers. Serf Master is a small python framework that aims to make Serf easier to manage.
https://github.com/garethr/serf-master
Anomaly detection was definitely one of the themes of lots of monitoring conversations this year. So crucible should be handy for people working in this area. It’s a small tool for testing detection algorithms, taking a time series data set, any algorithms you want to test and producing a simple visual output.
https://github.com/astanway/crucible
Docket is another tool intended for use with Docker, looking to move configuration and the docker images into version control with Git. Docket introduces a simple config file and has good getting started documentation and examples.
https://github.com/polydawn/docket
I’m quite a fan of code analysis tools but linking lots of them together and then using the output can be time consuming. Prospector looks interesting for Python folks, aiming to integrate pep8, pylint, pyflakes and complexity scores with an understanding of common frameworks like Django.
https://github.com/landscapeio/prospector