DEVOPS WEEKLY ISSUE #243 - 30th August 2015
One of the interesting things with being around the devops community for a bunch of years is watching and listening as common practices and themes emerge. Seeing another Devops State of the Union presentation in this issue brings home how the conversations are still ongoing.
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] 7 Practices to Expand Performance and Effective Collaboration in DevOps
When apps fail, whose fault is it? In today’s DevOps world, every stakeholder in the app delivery chain is accountable for various aspects of performance, scalability, and availability. In this webinar Mark Tomlinson, performance engineering veteran and founder of the popular PerfBytes podcast, will share seven practices to help you expand performance and effective collaboration into your DevOps team. Register now:
http://ow.ly/Rs67F
News
Postmortems are one of the core practices of many mature teams. This article explains in some details the what and why of running postmortems, with tips on how to run your own.
http://www.infoq.com/articles/postmortems-etsy
I’ve featured a few unikernel related posts in the last year and one of the topics I’ve been pondering is what operational challenges unikernels bring to the table. This post has some thoughts.
http://www.morethanseven.net/2015/08/21/operating-unikernel-challenges/
An update to the classic Devopsdays state of the union. Everything from high level meaning, via evolving patterns and practices to emerging technology.
http://www.slideshare.net/botchagalupe/devopsdays-chicago-state-of-the-union-2015
One of the practices that has emerged along with devops is writing tests for your infrastructure. Although common in professional software development writing automated tests might still be something you need to sell inside your organisation, this post is a good starting point.
http://blog.jetbrains.com/idea/2015/08/why-write-automated-tests/
A nice list of resources for systems administrators, mainly a list of tools categorised into everything from virtualisation to editors, but also includes a few book recommendations
https://github.com/kahun/awesome-sysadmin
Bosh, often used alongside Cloud Foundry, is super powerful but also quite a departure from similar tools and takes a bit of understanding. This simple tutorial is a nice place to start.
http://mariash.github.io/learn-bosh/
A personal option on why ITIL and ITSM practices fail. Somewhat one sided but some good observations in there, and the comments continue the discussion.
http://etherealmind.com/why-i-hate-itil-so-much/
A nice war story to accompany the above post on some of the potential failings of ITIL implementations. Optimising for change is definitely a part of adopting devops practices that’s often taken for granted in small organisations and probably the most important part in larger ones.
http://lkhill.com/war-stories-itil-process-vs-practice/
A video from the recent Euro Python conference focused on adopting a continuous delivery model for a typical Python application. Step by step examples using Devpi, Ansible and Jenkins.
https://ep2015.europython.eu/conference/talks/a-pythonic-approach-to-continuous-delivery
One of the strengths of Mesos is that you can bring your own framework if the existing ones (like Chronos or Marathon) don’t fit your usecase. Fenzo is a new scheduler from Netflix and the post explains why writing a new framework made sense.
http://techblog.netflix.com/2015/08/fenzo-oss-scheduler-for-apache-mesos.html
A presentation from the recent MesosCon all about using Mesos and Marathon to scale out a Jenkins installation.
https://speakerdeck.com/rji/distributed-ci-scaling-jenkins-on-mesos-and-marathon-mesoscon-2015
Jobs
Sainsbury’s Digital are looking for DevOps Engineers in London and Coventry. You’ll have a real bug for automation as well as a solid understanding of Linux. You’ll also be handy with a programming or scripting language, configuration management systems and common monitoring tools. This role is more than just a technical one, you’ll also be explaining concepts to non-technical colleagues and helping the rest of the business embrace DevOps.
https://http//www.sainsburys.work/dandt
Conde Nast are looking for a Senior DevOps Engineer to come and work on a new commerce start-up in London. We are looking for someone who knows how to deploy immutable instances or containers into distributed systems. We host our infrastructure in AWS, making heavy use of Terraform, Consul, and Chef.
http://bit.ly/1hCgUlC
Events
Icinga Camp is happening on the 10th of October, just after Puppet Conf, in Portland Oregon. Some of the talks have already been announced and you can register for the event too.
https://www.icinga.org/community/events/icinga-camp-portland/
Tools
Hyper is described as a hypervisor agnostic docker engine. It’s able to launch tiny vms using the docker interface, providing the isolation of hardware virtualisation with the speed and user interface of containers. Very interesting.
https://hyper.sh/
Portus is nice-looking user interface for the self-hosted Docker registry which also provides a team centric authorization service, so different teams can have there own namespace in the registry.
https://github.com/SUSE/Portus
Jungle is the start of a higher-level command line interface for AWS, currently supporting EC2 and ELB services.
https://github.com/achiku/jungle
Gotty is a handy application which exposes a command line application as a web application. It’s surprisingly configurable too.
https://github.com/yudai/gotty
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