Some posts this week on classic ops problems; like database migrations and runbooks - as well as some discussions of modern development workflows using containers.

Sponsor

Outages cost organizations millions. Grab this Forrester report, The Changing Landscape of IT Incident And Crisis Management, to see why a streamlined process for incident management needs to become top priority.
http://try.victorops.com/DevOpsWeekly/ForresterReport

News

A discussion of approaches to database migrations when adopting continuous deployment.
https://samsaffron.com/archive/2018/03/22/managing-db-schema-changes-without-downtime

A look at the lessons that lots of people took from ITIL that go against modern ITSM practices.
http://www.itskeptic.org/content/lessons-world-took-itil

A post comparing the progress of Google Cloud relative to AWS. With the growing popularity of multi-cloud strategies it’s useful to keep up-to-date.
https://medium.com/@jaychapel/google-cloud-platform-vs-aws-is-the-answer-obvious-maybe-not-1a11482beca9

A good post on the importance of runbooks for introducing development teams to operations tasks and increasing transparency.
https://blog.buildo.io/enable-your-devs-to-do-ops-9a0a870baa1

A post on the importance of making engineers accountable for behavior in production. Lots of good tips for those working in small teams around culture, security and minimising risk.
https://www.threatstack.com/blog/how-threat-stack-does-devops-part-iv-making-engineers-accountable/

A detailed post covering a modern software development workflow using Docker and Kubernetes - taking into account monitoring requirements, event buses, single-sign-on and more.
https://hackernoon.com/the-best-architecture-with-docker-and-kubernetes-myth-or-reality-77b4f8f3804d

security.txt is an approach to one of the problems of vulnerability disclosure, namely finding out who to disclosure issues to. It will be interesting to see if this gains traction.
https://securitytxt.org/

Java 9, and the upcoming Java 10 release, feature a number of improvements for running Java applications inside Docker. This post covers heap size, CPU availability and the attach API.
https://www.opsian.com/blog/java-on-docker/

A reminder that etcd, or similar services, are databases and should be treated with the same level of care when it comes to access control - lest you expose sensitive credentials over the network.
https://elweb.co/the-security-footgun-in-etcd/

To understand running Kubernetes in production it’s useful to have a local development workflow based around Kubernetes too. This post looks at Skaffold for continuous deployment to a local cluster.
https://nemethgergely.com/using-kubernetes-for-local-development/index.html

A quick introduction to containers, with a focus on what this means for Java and JVM-based applications. Includes details of JDK 9 integration with cgroups.
https://mesosphere.com/blog/java-container/

Jobs

We’re Nested, a London based startup taking the pain out of selling homes. We’re looking for a Platform Engineer to influence the development and deployment platform here. You’ll be an early member of our technical operations team and help shape our practices, platform and culture from the very beginning. Our tech stack currently includes Google Cloud Platform, Kubernetes, Terraform, PostgreSQL, Elixir, ReactJS, Python and more.
https://nested.workable.com/j/F69BC7758A

Tools

Teresa is an application deployment tool for teams, built atop Kubenetes. It provides a CLI tool for creating new teams, users and for deploying applications to the cluster.
https://github.com/luizalabs/teresa

Tern is a new tool aimed at providing information from containers that’s useful for compliance purposes. It’s currently focused on identifying the sources, versions and licenses of packages installed in container images.
https://github.com/vmware/tern
https://blogs.vmware.com/opensource/2018/03/13/open-source-tern-makes-containers-compliant/

Outages cost organizations millions. Grab this Forrester report, The Changing Landscape of IT Incident And Crisis Management, to see why a streamlined process for incident management needs to become top priority.
http://try.victorops.com/DevOpsWeekly/ForresterReport