DEVOPS WEEKLY ISSUE #110 - 10th February 2013
Logstash, one of the tools I’ve featured quite a few times in this newsletter has a book available, and a self published one at that. I’d love to see more self publishing around emerging tools and topics from the devops community. Traditional publishing is hard in such a fast moving environment, but being able to read a book rather than a blog post can really help with understanding a new tool or topic.
Sponsor
Devops Weekly is sponsored by Brightbox Cloud - launch cloud servers in multiple UK datacentres in seconds…
http://brightbox.com
News
The logstash book has been released! I read this coming back from FOSDEM and it’s well worth spending some time with. It covers why log management is important, how to get started with logstash and just as interestingly how to scale up and how to make decisions about the different logstash components.
http://www.logstashbook.com/
Platform as a Service architectures are interesting (at least to me), and this post, the first in a series, starts to dissect the inner workings of CloudFoudry. This post covers how the application is run with lots of code examples and by running one of the CloudFoundry components on it’s own.
http://starkandwayne.com/articles/2013/02/05/diy-paas-running-apps-with-a-dea/
Goo list of handy one-liners for knife, the chef command line tool. A mix of search tips and exec hacks, if you understand all of these you’ll probably be able to adapt them for lots of other uses too.
http://dougireton.com/blog/2013/02/03/knife-tricks/
An old post (from 2009) but worth reading amongst the recent spate of vulnerabilities. This covers timing attacks, and is probably the clearest (and scariest) description I’ve read of what’s quite an interesting attack vector for certain systems.
http://codahale.com/a-lesson-in-timing-attacks/
Even the skeptics are coming round to the Devops ideas it seems. Good post on how ITIL and other more traditional approaches can be seen in terms of Devops. Looking forward to more details.
http://www.itskeptic.org/content/unified-theory-itsm-reconciling-devops-and-itil
Interesting presentation from the recent Monkigras conference, on what the big data and data science community can learn from the devops movement. The phrase “analysis as code” should hopefully stick in the same way as infrastructure as code has done.
http://www.slideshare.net/dberkholz/data-science-devops-and-drinks-the-perfect-combination
Another survey, this one focusing on operations productivity. It’s 25 questions and the plan is to release the resultings in a month or so.
https://www.surveymonkey.com/s/productivity-survey
Some good tips on network tuning, nginx configuration and debugging performance issues around a simple but high throughput node.js application. Most of it should be relevant for anyone proxying to evented applications.
https://engineering.gosquared.com/optimising-nginx-node-js-and-networking-for-heavy-workloads
Jobs
Red Gate Software are looking for a DevOps Engineer to join their team. In the next year, Red Gate’s infrastructure will undergo more change than in the previous five. They’re committed to the cloud, to the blurring of the line between developer and sysadmin, to automation and the move towards infrastructure as code, and to monitoring as the foundation of responsive infrastructure. If you’re some peripatetic hybrid of developer and sysadmin, imbued with a depth of understanding of how this cool stuff can work for a mature company, then go and help them build it. Simply upload your CV and cover letter to apply.
http://www.red-gate.com/our-company/careers/current-opportunities/devops-engineer
Events
A Chef fundamentals training event is coming up on the 26th and 27th of February in Berlin. If you’re interested you can get a 75 Euro discount using the code CBCHEF-Q28.
http://de.amiando.com/cheftraining.html
Tools
Interesting plugin for Vagrant. This uses the ZFS filesystem to make snapshots of virtual machines, and claims to drastically improve the first boot time.
https://github.com/larstobi/vagrant-zfs
Monigusto is an integrated suite of monitoring tools, packaged as a series of Chef recipes and a vagrant configuration. It looks a great starting point if you want to experiment with a pre-configured monitoring stack.
https://github.com/monigusto