How do I set up replication from MySQL to MongoDB?

I have a bunch of data from a scientific experiment stored in a MySQL database, but I want to use MongoDB to take advantage of its map/reduce functionality to power some web charts. What is the best way to have new writes to MySQL replicate into Mongo? Some solution where I inspect the binary MyS开发者_如何学运维QL log and update accordingly, just like standard MySQL replication?

Thanks! Alex

---

MySQL and MongoDB uses very different data and query models, so you can't transfer data directly.

Alas, moving data between the two must be done manually, and doing that efficiently depends very much on your data. Eg. you could transfer each table to a separate collection (roughly a table in MongoDB-lingo), and making the unique attributes in the tables to the _id-attribute. Alternately, you can make the _id to tablename+unique_id.

Basically, as Document databases are essentially free-form, you are free to invent your on schemes ad-infinitum (as long as the _id-attributes are unique within the collection).

---

Tungsten Replicator is data replication engine for MySQL.

Using heterogeneous replication, you may be able to set up MySQL to MongoDB replication.

---

I am not familiar with MongoDB but my quick look shows it is incompatible with mySQL so unless someone has written something to import form mySQL you are out of luck.

You could write your own import function.

assuming your mySQL tables use an incrementing unique 'id' field you could track the last row in mysql and then send it to mongodb when it changes.

Alterations and deletions would be much more difficult to deal with. if this is important then inserting the data at the source is probably the best bet.

Do you need to insert the data into mysql at all? could you do it all in mongodb and save all the trouble?

DC