A range of content this week on cost-reduction for Kubernetes clusters, devops influenced software development lifecycle, some devops history, testing serverless applications and tips for data-driven post-mortems.

Sponsor

Embrace a DevOps culture of SRE and continuous deployment. Build reliability by establishing a system for constant reevaluation of technology, processes, and people.
http://try.victorops.com/DevOpsWeekly/SRE-book

News

An excellent post on the developer experience of testing serverless functions locally, looking at a real world example and demonstrating the AWS Serverless Application Model.
https://joaorosa.io/2018/05/21/life-of-a-c-developer-how-to-build-and-test-an-aws-lambda-locally/

A detailed look at how Netflix structure their development teams and implement a software delivery lifecycle. Good discussion of trade-offs and advantages with developers operating their own services in production.
https://medium.com/netflix-techblog/full-cycle-developers-at-netflix-a08c31f83249

Slides from my talk at Agile Portugal last week, a bit of history of the devops movement as well as an overview of topics being discussed in the community now and (maybe) next.
https://speakerdeck.com/garethr/devops-the-continuing-evolution-of-operations

A post on one of the challenges facing the Kubernetes community, that of complexity, and more interestingly where the end-user experience should lie.
https://www.influxdata.com/blog/will-kubernetes-collapse-under-the-weight-of-its-complexity/

By making deployment and other systems administration tasks easier the conversation about what is left for operators sometimes comes up. This post explores the question and makes several good points.
https://www.rundeck.com/blog/reports-of-the-impending-demise-of-ops-are-greatly-exaggerated

A nice workflow for packaging up tools as docker images, in this case to use them on Windows via NPM tasks.
https://blog.stephencleary.com/2018/05/docker-as-a-tool-provider.html

Good tips for reducing the cost of running Kubernetes clusters, some AWS specifics. Covers choosing the correct instance size and type as well as looking at vertical and horizontal autoscalers.
https://medium.com/@dyachuk/why-do-kubernetes-clusters-in-aws-cost-more-than-they-should-fa510c1964c6

A presentation containing some tips for running data-driven post mortems for incidents, including a full example of a real-world incident.
https://drive.google.com/file/d/1i7hrTGtk_abE1-L68Mr6VH5V6LwjQuYD/view

Sponsored

O’Reilly Velocity Conference | June 11-14 in San Jose, CA Velocity is just two weeks away! Don’t miss your chance to get practical training in microservices, cloud migration, container orchestration, DevOps, performance, security, and much more. Check out the full program to see how we can help you make your systems fast, reliable, and secure. Passes start at $199. Register today and save 20% on your Gold, Silver, or Bronze pass using code DOW20!
https://oreil.ly/2GZ2ysI

Tools

A handy tool for forensic analysis of Docker containers. Docker Explorer provides utilities for exploring the layered file system at different points.
https://github.com/google/docker-explorer/

Embrace a DevOps culture of SRE and continuous deployment. Build reliability by establishing a system for constant reevaluation of technology, processes, and people.
http://try.victorops.com/DevOpsWeekly/SRE-book