Accelerate Delivery Newsletter #9
#9 — OCT. 31, 2018
Accelerate Delivery Newsletter

Greetings, ADers!

"The error happens intermittently..." 😱🎃 Happy Halloween to our subscribers in the US. Keep reading to learn how CD can help you avoid production horror stories.

Bill - Accelerate.Delivery team

Your chipper shipper

Continuous Delivery For Hiring

Continuous Delivery removes the friction of delivering code to production. We can apply similar principles to the hiring process in software engineering. Let's examine some experiments you can run on your own hiring pipeline and the benefits you should expect to achieve.

Accelerate Delivery

Hands-On Canary Deployments with Istio and Kubernetes

This post provides a step-by-step guide to setup canary deployments with Istio and k8s. This is a great intro to Istio if you're already familiar with k8s.

GitHub's October 21 post-incident analysis

Postmortems are retrospective events in continuous delivery. They are useful to calculate Mean Time to Recovery (GitHub's was 24 hours and 11 minutes for this event). They are also opportunities to see what can be improved. A postmortem also reinforces trust. This best-of-the-business postmortem goes into incredible detail for an outage event at GitHub's data centers.

The GitHub Blog

🔊 Podcasts

DevSecOps with Edward Thomson

Gets right to the point.

How would you change your continuous integration pipeline to acknowledge security?

Listen for actionable steps you can take to integrate security checks into your delivery pipeline.

Software Engineering Daily

🖥️ Videos

The Data Behind DevOps: Becoming a High Performer

Discuss the measurements that are antipatterns and the measurements that are worth it. Antipatterns include: Lines of code, Velocity, Utilization. Nicole and Jez encourage us to focus on:

Outcomes not outputs. Global measures, not just local measures.

The measurements discussed in the video encourage us to optimize on speed and stability. Both are possible without trade offs.

YouTube - DevOps Enterprise Summit

Increasing Deployment Safety with Metric Based Canaries

Brandon from Lookout talks about the transition from continuous delivery to continuous deployment. Canary deployments with automatic monitoring enabled this shift. Brandon's key lessons about canaries are:

  • It takes time to dial in what metrics to watch
  • Too many metrics will cause noise and false failures
  • Slows down deployments by hours, but saves time

In the near future Brandon will keep an eye on Kayenta, a Spinnakr integration to monitor canary deployments.