DEVOPS WEEKLY ISSUE #267 - 7th February 2016
A busy last week, thanks to everyone who helped run, spoke at or attended Configuration Management Camp in Gent. Lots of content from the conference below and I’m sure I’ll find more next week too.
Sponsor
Devops Weekly is sponsored by Brightbox Cloud: the multi-zone cloud platform built for High Availability.
Try an SSD cloud server in 30 seconds with your £20 free credit…
http://brightbox.com/devopsweekly
Sponsored
[eBook] 7 Performance Metrics to Release Better Software, Faster
There are 7 key metrics you must know to deploy software faster without failing faster. Read this eBook to learn these, with real-life examples, so you can avoid problems later.
http://ow.ly/U3BJ7
Configuration management Camp
My own talk on the second day covered building APIs for container runtime and the importance of metadata, and metadata standards, in managing containers in production.
https://speakerdeck.com/garethr/managing-container-configuration-with-metadata
A nice talk about what orchestration is from a higher level, and where it fits in with other conversations around configuration management, scheduling and automation in general.
https://speakerdeck.com/liamjbennett/everything-is-an-orchestration-problem
Orchestration was definitely a topic of discussion this year. Another talk looking at constraints (using the example of the New York subway) and the relationship to model-based configuration management problems.
http://www.slideshare.net/lutter/orchestration-and-the-new-york-subway-57828762
Good talk from the event about using automation to build self-service systems, in this case empowering developers to deploy their own data stores.
http://www.slideshare.net/bobtfish/empowering-developers-to-deploy-their-own-data-stores
Quick introduction to the InSpec testing framework, with examples of code alongside a discussion of the importance of demonstrating compliance in code, rather than spreadsheets.
http://www.slideshare.net/mpgoetz/inspec-or-how-to-translate-compliance-spreadsheets-into-code
A look at how the OpenStack infrastructure group use Puppet and Ansible together to manage infrastructure. Lots of code examples and a story of how this has evolved over time.
http://spencerkrum.com/talks/better-together-cfgmgmtcamp2016/slides/#1
Comprehensive raw notes from various talks at configuration management camp.
https://gist.github.com/philandstuff/7fe6037ad2c1fcdab999
Interesting look at the the reasons for having a GUI for a configuration management tool (in this case Salt) and an introduction to the new Saltpad application.
https://speakerdeck.com/lothiraldan/saltpad-a-saltstack-web-gui
News
Especially if your audience is a technical one details in a public incident report go someway to making up for any disruption. This is an excellent example, with a good learning points, regarding a recent GitHub outage.
https://github.com/blog/2106-january-28th-incident-report
I tend to avoid industry news style posts but this one makes some great points about the future direction of containers, unikernels and VM based infrastructure with regards the user interface. Worth reading.
http://it20.info/2016/01/why-docker-containers-and-docker-oss-docker-inc/
Two posts about operating Elasticsearch, covering choosing it in the first place through monitoring and scaling to hundreds of millions of documents.
https://signalfx.com/how-we-monitor-and-run-elasticsearch-at-scale/
https://signalfx.com/scaling-elasticsearch-sharding-availability-hundreds-millions-documents/
A good post arguing that we need a new approach to systems development where “continuous partial failure” is the normal state of affairs. A few examples and more posts to come in the series.
http://blog.cognitect.com/blog/2016/2/3/the-new-normal-failure-is-a-good-thing
Some interesting, if obviously skewed, data on container adoption from the Kubernetes community. The challenges are worth reading for anyone considering containers at the moment.
http://blog.kubernetes.io/2016/02/state-of-container-world-january-2016.html
Raising the profile of a technology tends to bring about questions and critiques, which can be useful in improving or validating that technology. Several recent posts have talked about potential issues with Unikernel based systems - this one focuses on the single address space, management and complexity.
http://etbe.coker.com.au/2016/02/04/unikernels/
Jobs
Looking for new DevOps opportunities? Spend your time building, not searching for a job. Sign-up for Hired today.
http://hrd.cm/23r2ok5
Events
The first Sheffield Devops is next week, on Thursday the 11th at Union Street. A new community meetup with two talks (An Introduction to DevOps in Microsoft Azure and Everything you ever wanted to know about Devops but were afraid to ask) and plenty of room for discussion.
http://www.sheffielddevops.org.uk/?og=1
Tools
Simple tool for talking the contents of a configuration file and saving it as a Kubernetes secret for use by other Kubernetes resources.
https://github.com/kelseyhightower/conf2kube
Uphold is an interesting looking tool intended to verify backups. It downloads the backup, decompressing, loading and then running programmatic tests against it.
https://github.com/forward3d/uphold
Devops Weekly is sponsored by Brightbox Cloud: the multi-zone cloud platform built for High Availability.
Try an SSD cloud server in 30 seconds with your £20 free credit…
http://brightbox.com/devopsweekly