Meteor integration with enterprise data sources e.g. Oracle, Hadoop/HDFS/HIVE etc.

[Alice Smith]

Hello

Are there any plans/timelines for integration of Meteor with enterprise data sources e.g. Oracle, Hadoop/HDFS/HIVE etc. ?

If so, can somebody share a rough timeline? Are there any current efforts (maybe in alpha/beta phase) or any workarounds (RESTful APIs, Webservices etc.)?

Thanks.

[Abigail Watson]

If the data source can be exposed through an HTTP interface or JSON stream it can probably already be integrated.  The real problem is with object-relation mapping (ORM) layers, which Meteor doesn't support.  Keep in mind that Meteor's performance and functionality is very much related to not having the overhead of an ORM layer, so it's something that people are reluctant to add in.  But if the enterprise system can take care of the ORM mapping and generate a stream of JSON objects, Meteor can consume that a couple of different ways (i.e. directly into the Mongo database, and use Mongo as a persistent object store/caching layer, or directly to minimongo on the client side.

[Alice Smith]

Okay, thanks Abigail.

We have decided to use JPA entities exposed through RESTFul interface (as JSON streams) that Meteor can consume. 

So far, it has worked with Oracle and Hadoop/Hive (with JDBC connectors). Performance is almost as good as MongoDB, so that is very encouraging. But we are testing extensively in a stage environment.

-Alice

[Serkan Durusoy]

Hi Alice, I suggest you take a look at http://restlet.com/products/restlet-framework/ for light-speed rest interface development