Hbase and EC2 for Google-like Search

I have built a SaaS company based on PHP and MySQL running on cPanel accounts with each customer running their own separate MySQL database. At some point I would like to create various applications for searching of data across customers. That's not possible right now because each customer is using a separate database.

I just discovered Hbase. It's a distributed database built on top of Hadoop, an Apache project for reliable, scalable, distributed computing. Hbase was modeled after Google's Bigtable distributed storage system which they use to support web indexing, among other Google projects. Hbase's goal is to accomodate very big tables, with billions of rows and millions of columns hosted on a cluster of servers running Hadoop. Check out the Hbase wiki. Hbase is already powering several websites including Mahalo and VideoSurf. Here's the full powered-by list.

Lacking any experience in distributed computing, I would need commercial assistance to implement Hadoop/Hbase. Startup Cloudera provides commercial support for Hadoop. Cloudera has an impressive team and support the deployment of their Hadoop distribution on Amazon EC2 servers.

I have a lot to learn about Hadoop to see if this will be a viable solution for me. My next step will be to watch Cloudera's Basic Hadoop Training videos.