DEVOPS WEEKLY ISSUE #149 - 10th November 2013
A busy week coming up, with Devops Days London on Monday and Tuesday, followed by Velocity in London for the rest of the week. If anyone is around at either of these do come and say Hi.
Sponsor
Devops Weekly is sponsored by Brightbox Cloud - launch cloud servers in multiple UK datacentres in seconds…
http://brightbox.com
News
A good getting started guide for Marelle, the prolog based configuration management system. Probably mainly interesting for the discussion of logic programming and how it maps quite nicely to various automation tasks.
http://quietlyamused.org/blog/2013/11/09/marelle-for-devops/
A light-hearted look at what a whole range of science fiction material can teach us about the importance of collaboration between developers and operations.
http://www.slideshare.net/geekygirldawn/sci-fianddevopsignite
Nice writeup of how PagerDuty use Chef, but focusing on the people and environments rather than code. Discusses some of the tools used to support Chef, as well as approaches to testing infrastructure concerns.
http://blog.pagerduty.com/2013/11/chef-at-pagerduty/
The start of what looks like a useful series reviewing various dashboarding tools. The first review is for Tasseo for Graphite and features a good description of the pros and cons of using it. Hopefully more reviews soon.
http://dashboarddude.com/blog/2013/11/09/dashboard-tasseo/
The call for proposals is out for the testing and automation dev room at FOSDEM in Brussels next February. Last year was excellent and this year the organisers are looking for 30-40 minute talks and 10 minute lightning talks, to be submitted by the 30th of November.
http://fosdem-devroom.org/wiki/Testing_and_Automation_(2014)
A good introduction to the InfluxDB time-series database, covering the core featureset, the basic operations using the ruby client and future direction of the project.
https://speakerdeck.com/auxesis/influxdb-making-time-series-fun
Facebook have released Presto, a distributed SQL engine for querying huge datasets in HDFS aimed at being used for low latency queries. The post covers lots of the engineering details and the support for standard ANSI SQL is impressive.
https://www.facebook.com/notes/facebook-engineering/presto-interacting-with-petabytes-of-data-at-facebook/10151786197628920
http://prestodb.io/
A nice writeup of getting started with Statsd, with examples of using Graphite and Librato to send the metrics to. Also explains how this stack sit alongside other monitoring tools.
http://red-badger.com/blog/2013/11/04/fast-and-flexible-monitoring-with-statsd-graphite-and-librato/
Jobs
TechCity based start-up Streamhub are looking to leverage the power of DevOps in order to support the rapid growth of their new media analytics platform. If you have experience in big-data technologies, are creative in your thinking and interested in joining a award winning team, please visit their website http://www.streamhub.co.uk/jobs.php or email your CV to [email protected].
Tools
Rearview is a very interesting self-service monitoring tool, allowing monitors to be easily created against data in graphite from within a nice looking web interface. Alerts can then be routed out to PagerDuty or other systems, with the monitors themselves written in-line in Ruby.
https://github.com/livingsocial/rearview
InfluxDB is an open-source, distributed, time series, events, and metrics database written in Go. It features a nice SQL-like API for querying metrics and aims to make clustering and horizontal scaling simple.
http://influxdb.org/
Chef Taste is a nice looking command line tool that lists out the dependencies for a Chef cookbook, along with whether they are up to date or not. A nice touch is a link to the changelog for out of date dependencies.
http://www.arangamani.net/blog/tech/2013/10/27/introducing-chef-taste-to-taste-your-cookbooks/