DEVOPS WEEKLY ISSUE #146 - 20th October 2013
Devops weekly now has over 8000 subscribers, with more signing up each week and with more and more mainstream articles, whitepapers or conferences talking about devops. But most of the best conversations are still at small events, on personal blog posts or freely given away by engineers in large companies. I just want to say thank-you to that community of people publishing high quality content.
Sponsor
Devops Weekly is sponsored by Brightbox Cloud - launch cloud servers in multiple UK datacentres in seconds…
http://brightbox.com
News
A good story about one organisations problem to provision a new virtual machine in less than 6 weeks. A common problem in some organisations, the post also has some realistic suggestions for how to fix problems like this.
http://kief.com/change-management-fail.html
Setting up SSL, and importantly getting the configuration just right, can be tricky. This guide from Mozilla contains specific confirmation recommendations and information about the different features and the risks they mitigate.
https://wiki.mozilla.org/Security/Server_Side_TLS
An interesting addition to the branch by abstraction and feature toggle approach to adding new functionality to a code base. Suggests an internal verification process to check both paths work as expected.
http://www.stephen-smith.co.uk/application-pattern-verify-branch-by-abstraction/
Interesting presentation about how Rackspace are using Ansible to standardise various existing ad-hoc automation efforts and replace various existing shell script based solutions.
http://www.slideshare.net/JesseKeating/ansiblefest-rax
A good introduction to devops slide deck which focuses on the story of one particular project. The focus on sharing, interaction and culture looks great - I’ll look out for the video when it’s released.
http://www.slideshare.net/thekua/devops-from-the-ground-up
A solid reading list of books, blogs, videos, podcasts and conferences on the subject of technical operations. A good starting point if you’re just getting into the discipline.
http://kovyrin.net/2013/09/23/techops-resources/
An article about implementing continuous delivery. Makes good points about going from zero to a complete pipeline can seem daunting, but the individual small steps can add value quickly.
http://www.drdobbs.com/architecture-and-design/continuous-delivery-the-first-steps/240161356
A nice technical comparison of the different approaches taken by Facebook and Twitter for building their internal analytics platforms. Good comparison of a logging vs event driven approach.
http://natishalom.typepad.com/nati_shaloms_blog/2013/10/facebooks-vs-twitters-approach-to-real-time-analytics.html
Tools
PKGR had some nice updates last week that in my view make this potentially hugely useful. PKGR combines Heroku buildpacks and system packages to provide a very nice interface for building system packages of your applications.
https://github.com/crohr/pkgr
Ansible has a number of modules built-in which allow you to use it for managing virtual machines from various cloud providers. This project demonstrates a simple integration with Digital Ocean, creating an entire infrastructure from a simple YAML file.
https://github.com/garethr/ansible-provisioner
DNSYO is a little tool built to help keep track of DNS propagation. In short, it’s nslookup, if nslookup queried over 1500 servers and collated their results.
http://samarudge.github.io/dnsyo/
Megaphone looks like an interesting little monitoring tool. It’s designed to be run locally on all your machines, and provides a HTTP and JSON based interface to checks, which should make integration with other tools easy.
https://github.com/looprock/Megaphone