> Generating hash codes is not a simple or quick operation when you are
> considering it in relation of a heavily used transaction processor.
Nope, that's premature optimization.
I'm sure the Journal system will introduce new performance bottlenecks, but hashing is not one of them. Even my 2-year-old CPU (Phenom II) can hash ~300 MB/s of data per second with SHA-512, on a single CPU core. SHA-1 and MD5 are even faster.