DEVOPS WEEKLY ISSUE #379 - 1st April 2018
Running the gamut of topics this week with some great posts on technical support, centralising traditional operations tasks, deployment tooling, the importance of scripting, immutable infrastructure, secrets management and more.
Sponsor
Chaos engineering is the future of continuous integration and development. Creating a DevOps culture of reliability begins with developing an SRE structure of collaboration and accountability. Read how VictorOps is using “Game Days” to create a strong SRE culture.
http://try.victorops.com/DevOpsWeekly/GameDays
News
A series of posts on evolving a deployment process. Starting with a fairly typical automated SSH/Git based process and stepping through the design decisions for adopting container-based tooling.
https://medium.com/@AaronKalair/all-aboard-the-deploy-train-stop-1-anyone-for-tea-a5c12b984ed9
A walkthrough of why Helm is important to the Kubernetes ecosystem, covering what problems a package manager sets out to solve and a discussion of some of the related tooling.
https://platform9.com/blog/kubernetes-helm-why-it-matters/
A good post on the tension between pushing operational responsibility out to the development teams, and the benefits of centralising certain aspects of ops, especially those related to traditional service management.
https://aws.amazon.com/blogs/enterprise-strategy/the-critical-missing-piece-of-devopsand-how-to-find-it/
A post on the importance of generative tooling and scripting for administrators, and in this specific case why PowerShell is a required skill for Office365 administrators.
https://www.petri.com/powershell-important-office365-admins
Brigade allows for scripting a set of steps to run in containers. This post explores how to share data as part of your pipeline, looking at both the pipe concept and the use of a shared filesystem, and what this looks like in the API.
http://technosophos.com/2018/03/26/how-brigade-shares-data-between-containers.html
Investigating swarming as an approach to technical support, and a look at applying swarming to deliver cynefin.
https://medium.com/@JonHall_/using-swarming-to-deliver-cynefin-in-tech-support-34dc4992e3e0
Instructions for configuring hiera to lookup secrets from Vault in your Puppet runs, including some useful tips for improving performance with hiera 5.
https://petersouter.xyz/how-to-use-vault-with-hiera-5-for-secret-management-with-puppet/
A good overview of the various options available when running Java applications in Docker containers - looking at the various operating system base images and different flavours of JDK.
https://medium.com/oracledevs/running-spring-boot-in-a-docker-container-on-openjdk-oracle-jdk-zulu-on-alpine-linux-oracle-c9dd56b75f76
A look at configuring the Drone CI/CD system to run on Kubernetes. Some interesting container-native integration bits, including Helm and Docker support in the pipeline definition.
https://akomljen.com/set-up-a-drone-ci-cd-pipeline-with-kubernetes/
A nice introduction to monitoring applications with Prometheus. Including starting your own instance, examples of instrumentation and querying for metrics using PromSQL.
https://blog.alexellis.io/prometheus-monitoring/
A slide desk stepping through immutable infrastructure, explaining how it varies from other approaches which rely on mutable systems and what the benefits of this approach are.
https://www.slideshare.net/rhirschfeld/srecon-immutable-infrastructure-rethinking-configuration-mgmt
Tools
Netassert is a network testing tool which supports Kubernetes as well as plain networks. You describe assertions in YAML, and the README contains lots of examples to help getting started.
https://github.com/controlplaneio/netassert
Building a hello-world style Lambda function is relatively straight-forward, but what does that look like when expanded to a more featureful application? This repository has a nice examples in Go which has lots of good examples.
https://github.com/nzoschke/gofaas
Dashbling is a handy looking dashboard application. It’s intended to be easy to confure with your own widgets, which can be shared as NPM packages.
https://github.com/pascalw/dashbling
Chaos engineering is the future of continuous integration and development. Creating a DevOps culture of reliability begins with developing an SRE structure of collaboration and accountability. Read how VictorOps is using “Game Days” to create a strong SRE culture.
http://try.victorops.com/DevOpsWeekly/GameDays