From the monthly archives:

December 2011

Chris Mattmann is a Software Engineer at NASA’s JPL. He’s the VP of OODT in the Apache Software Foundation and an adjunct professor at USC.

OODT is a framework for managing data from multiple sources and adding them to other data sources for different purposes (like a database and a search engine.) It manages hundreds of thousands of job in a day and terabytes or petabytes of data from various sources.

Mentioned in this episode:

  • Apache OODT
  • Nutch
  • Hadoop
  • Apache Software Foundation
  • NASA
  • NASA JPL
  • ftp
  • sftp
  • Solr
  • Lucene
  • Hive
  • File Catalog vs Search Engine
  • Tika
  • Goodle
  • Project Management was the hard part
  • Assume that failure happens and recover quickly
  • Ganglia
  • Torque
  • PBS
  • struts
  • IDL
  • CHLA (Childrens Hospital of LA)
  • VPICU
  • OODT Contact page (info on mailing lists, etc.)

{ 0 comments }

Today I am joined by Karl Wright, Nokia engineer, ManifoldCF developer and author of ManifoldCF in action. We discuss ManifoldCF, an Apache Incubator project, its beginnings, its purpose and its inner workings.

{ 0 comments }

When preparing to add some layout features to JotRod, I realized it was a Rails 3.0.9 application. Here’s a quick rundown on upgrading to Rails 3.1.3.

Download 164.7 MB
Download (iPod & iPhone) 37.9 MB
Take the 2011 Readers Survey

{ 5 comments }

AppDynamics is a company that provides a monitoring solution for .NET and Java platforms. I spoke with the VP of Engineering and one of the developers of the AppDynamics platform to dig into how they instrument your Java or .NET code and some of the tricks for following transactions from beginning to end.

There were a lot of neat tricks in this podcast episode

{ 0 comments }

In order to get someone a timeline in JotRod, we need followers and following lists to compile the Jots from. This means that we need to add a new ColumnFamily called Followers and another one called Following. We don’t have the joins capability from relational databases to do this for us.

I’m going to hijack the User model’s database connection to create the ColumnFamilies. (We don’t have migrations, yet.) Here’s what I ran in the rails console:

cf_def = CassandraThrift::CfDef.new(:keyspace => "JotRod", :name => "Followers")
User.connection.add_column_family(cf_def)
cf_def = CassandraThrift::CfDef.new(:keyspace => "JotRod", :name => "Following")
User.connection.add_column_family(cf_def)

Now that we have the ColumnFamilies, I want to have syntax like this to define the relationships on the User model:

list :followers, :User
list :following, :User

This should provide the following API:

#followers – returns an array of users as specified from the Followers ColumnFamily
#followers<<(user)  – adds the user to the User object’s followers list if it’s not already there

and a similar API for following.

Sandra’s repository

JotRod’s repository

Download 680 MB
Download (iPod & iPhone) 165 MB
Take the 2011 Readers Survey

{ 2 comments }