That is an Avro schema embedded in ruby code. Perhaps I should put it in a separate entry?
That is a bug, re: missing user_id. I will get a working github project out that will be bug free(ish) soon. Russell Jurney twitter.com/rjurney [email protected] datasyndrome.com On Dec 7, 2011, at 8:28 PM, Ayon Sinha <[email protected]> wrote: > Thanks for this. A few questions: > Is this Ruby code? > require 'rubygems' > require 'avro' SCHEMA = <<-JSON > { "type": "record", "name": "Email", "fields" : [ {"name": "message_id", > "type": "int"}, {"name": "topic", "type": "string"}, {"name": "user_id", > "type": "int"} ]} > JSON > And what happened to user_id (maybe a typo?) > grunt> DESCRIBE messages > avros: {message_id: int,topic: chararray} > > -Ayon > See My Photos on Flickr > Also check out my Blog for answers to commonly asked questions. > > > > ________________________________ > From: Russell Jurney <[email protected]> > To: [email protected]; [email protected] > Sent: Wednesday, December 7, 2011 2:40 PM > Subject: Booting the analytics application: events -> ruby -> avro -> pig -> > voldemort -> sinatra -> web browser -> user > > I wrote up a pig-centric tutorial about how to get started collecting data > with Avro, processing it with Pig and publishing it with Voldemort/Sinatra: > http://datasyndrome.com/post/13707537045/booting-the-analytics-application-events-ruby > > I hope it may be helpful for some. There is actually no Hadoop in the > intro, as it wasn't neccessary to get started using the tools to build > things - albeit on smaller data ;) Still, its nice to pick a set of tools > and processes that will scale to hell and back, and then not worry about > switching platforms later. > > -- > Russell Jurney > twitter.com/rjurney > [email protected] > datasyndrome.com
