Hello forum,
I'm in need of object storage for Mahout. BDB would be perfect but their
license is not compatible with ASF.
So I'm thinking I should clone your HashMap, make all access to element
data abstract and run it on ByteBuffers.
It would be used as a Map<K, DataFileEntry> pointing at where in the
object data file the current value is located.
Updaing values would mean to mark the old instance deleted, add a new
one to the end of the object data file and update the position in the index.
It would use Hadoops Writable to serialize keys and values.
It would initially be transactionless.
Any comments to this? Perhaps something similar already exists ASLed?
karl
