I was thinking of mentioning Hadoop/MapReduce or Cassandra, but these technologies are a big change from Radu’s current implementation, and the learning curve is steep.
Very cool technology though. At my work I design/maintain a 500 gigabyte SQL Server data warehouse with almost 400 tables in a denormalised star schema – have been trying to get my bosses to look at hadoop and MapReduce, but it’s ‘new’ tech and therefore complex and scary 🙂
