Hi Ashish, Sorry for the late response. It's a tough question :-) Especially for me because I am biased towards MetaModel and I don't know all the details of Drill...
But my 5 cent wrt. what the main difference is (from helicopter view): It seems that Drill is aiming to expose Hadoop and NoSQL in a JDBC-compliant interface so that existing application can use those datastore types in e.g. a BI reporting tool. Their query engine is thus built to be JDBC compliant and work with the string-based queries that you normally use when interacting with JDBC. MetaModel has a slight different approach to the "multitude of datastores" issue. JDBC is not considered the end-goal, but rather just one of the many types of datastores you will see. Rather we are trying with MetaModel to build a new "defacto standard" that will allow you to use the same query API (which is object-oriented and not (preferably) String-based) for many datastores. It also seems that we are emphasizing less on Big Data in the MetaModel space and more on "all kinds of data". For me personally the most important modules in the MetaModel are the file-based modules like the CSV module and the Excel module, because those file formats normally don't have any good query engines and because those are the formats that we want to treat the same way as we want to treat a proper database. Best regards, Kasper 2015-03-25 18:08 GMT+01:00 Ashish Mukherjee <[email protected]>: > Hello, > > I have been evaluating MetaModel as a possible component for my > architecture stack. > > Was wondering if anyone has explored Apache Drill too, which seems to bear > some similarity (SQL interface to multiple data stores). Any major pros and > cons of one against the other? I have been playing with MetaModel lately, > but new to Drill. > > Regards, > Ashish >
