I spent some time this week at the excellent ThingMonk IoT conference (thanks Fintan and James). One of the things that struck me was the overlap with the topics normally covered in this newsletter - especially around metrics and monitoring, but also some interesting new twists - like physical security, distributed device management and more focus on power. If anyone comes across, or is working on, writing on IoT system administration do drop me a line.

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

[Webinar] How to Build a Metrics-optimized Software Delivery Pipeline

Learn a practical step-by-step approach to optimizing your delivery process so you can deploy better quality software faster! You’ll learn why you should move to a metric-driven pipeline, which key quality metrics to measure and how to integrate them to catch problems earlier.
http://ow.ly/UTMwE

News

An excellent getting started with distributed systems post, covering basic introductions to failure, leader election, time and more. Includes a good set of links to papers as well.
http://videlalvaro.github.io/2015/12/learning-about-distributed-systems.html

A look at how GitHub manage their physical servers, and the tools they’ve build to make this low touch. Includes examples of integration with a chatbot and an internal inventory service.
http://githubengineering.com/githubs-metal-cloud/

A great detailed post answering the question “You login to a Linux server with a performance issue: what do you check in the first minute?” 10 commands from uptime to top via sar, vmstat and more.
http://techblog.netflix.com/2015/11/linux-performance-analysis-in-60s.html

Adopting devops practices often leads to reorganising teams and organisations. This presentation is a great introduction to the often feared reorg. Some good high level advice on what to expect and what is required for success.
https://speakerdeck.com/skamille/reorgs-happen

A nice introduction to layers in Docker images, and some tricks to take advantage of them. Also introduces a simple browser based tool for visualizing layers for an image.
https://labs.ctl.io/imagelayers-io-docker-visualization-and-badges/

The first post in a series covering using Elasticsearch as part of a time series monitoring setup. The post explains the different alternatives and why the author is experimenting with Elasticsearch.
http://engineering.laterooms.com/elasticsearch-as-a-time-series-database-does-it-work-part-1/

Some good numbers from a well respected research firm on the current adoption and trends around container and microservice adoption, with pointers to real-world case studies.
http://www.slideshare.net/dberkholz/devops-containers-microservices-separating-the-hype-from-the-reality

A deep dive into building an application on AWS Lambda, with lots of more general points about serverless architectures.
https://medium.com/in-pursuit-of-serverless-architecture/what-if-we-didn-t-need-an-app-server-anymore-f5b6586c58f4

Not a new post, but I don’t think I’ve included it before. A great write-up of why code freezes, although logical, often lead to more instability in well managed systems. “Shipping code is not the only change in a system”.
http://www.xaprb.com/blog/2014/11/29/code-freezes-dont-prevent-outages/

Events

A short online course focused on practical Devops related tools; from Vagrant and Jenkins to Docker and Kubernetes. The code DEVOPSWEEKLY provides a nice 50% discount for readers of the newsletter.
http://devops1.teachable.com/courses/learn-devops-continuously-deliver-better-software?product_id=47762&coupon_code=DEVOPSWEEKLY

Tools

Snap is a new tool for gathering of telemetry from systems, with a focus on being particularly modular and easy to administer. The accompanying blog post goes into some of the why and where the project is going.
https://github.com/intelsdi-x/snap
http://nickapedia.com/2015/12/02/what-if-collecting-data-center-telemetry-was-a-snap/

There are always risks with a rapid shift in technology, and in normally slow moving organisations adopting fast moving technology. This cost looks a containers and picks out several concerning trends or gaps in understanding which could lead to problems.
http://www.juliandunn.net/2015/12/04/the-oncoming-train-of-enterprise-container-deployments/

Documentation is an important part of any internal or external application. Slate is a documentation generator focused on API documentation. It provides functional and attractive templates and works well on mobile too.
https://github.com/tripit/slate

Herokuish is a tool for bringing parts of the heroku workflow to containers. Specifically it allows for use of buildpacks and the Procfile with Docker, and aims to be a low level component for others to reuse that require this functionality.
https://github.com/gliderlabs/herokuish

The Docker event stream is underappreciated in my experience, giving you a stream of everything happening in Docker. Ahab is an interesting new tool for processing this stream in Python, the blog post gives an example of using it for dynamic assignment of IP addresses.
https://tech.instacart.com/ahab/
https://github.com/instacart/ahab

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