I wonder what motivation they had for developing their own BigTable/Dynamo implementation rather than going with Cassandra or HBase.
Now there are three Apache projects with quite a bit of overlap (distributed columnar/key-value database). Hooray for fragmentation!