Persisting with MySql and Replicating to Mong/Couch

Question

Hi , I'm trying to move a reporting engine from a MySql DB (which is also the one used for batch jobs).

while we stil need the benefits of an ACID DB for batch processes we have no need for them in reporting also map-reduce can benefit reporting much .

do any of the two above have a way in which they can read the master logs (of mysql) and upload to data to himself ? (without me writing that code) .

also is anyone here using a map reduce db for reporting ? would be happy to hear his/her notes .

thanks