In time, Drill will likely be able to satisfy these requirements. The only thing that isn't directly in sightline is a large amount of analytical requests per second. Initially, we're really focused on queries between 100ms and ~20min. If you're looking for queries in the few ms range, we've generally found that, for now, storage layer specific clients are your best option 10's of thousands requests per second.
J On Mon, May 13, 2013 at 4:02 PM, Andun Sameera <[email protected]> wrote: > Hi All, > > These days me and 3 other batch mates of my university are doing a research > project to implement a open source perception analysis platform. Basically > it is most equal to a social network. In that we have following > requirements, > > - A big data store which is capable of providing high spreed querying. > Thus we choose a polyglot data tore built on top of HBase + MongoDB. > - A analytical tool which do on demand high speed analytics - When a > request comes to do analytics, tool should do the analytic quickly and > return the result. > - The above analytical tool should facilitate programmatic creation and > execution of analytical queries - All the execution and creation of > analytical quires are done though JAVA code or any other appropriate of > programming language. > - The amount of analytical requests per second is pretty high. > > I have read about Dremel and Apache Drill's capabilities. My question is > can I fulfill my above highlighted requirements using Apache Drill? > > Thanks! > -- > Regards > Andun S.L. Gunawardana > Undergraduate > Department of Computer Science And Engineering > University of Moratuwa > Sri Lanka > > Blog - http://www.insightforfuture.blogspot.com/ > LinkedIn - http://www.linkedin.com/pub/andun-s-l-gunawardana/34/646/703 > Twitter -http://twitter.com/AndunSLG
