Hi Rob. 1) impala is i think still an HDFS specific technology, and a cloudera centered one as well. i know the MapR folks have tried to broaden it some in their own fork,but that didnt get merged into impala head.
2) also, like cos said - im also willing to encourage anyone to maintain impala in bigtop, but i dont think the people who are right now most actively contributing to bigtop, are really involved w/ impala at all. I think focusing on spark, tez and emerging SQL workloads around those, will be more likely to be easily maintained in bigtop. but i have no objection either ! that said - if someone is going to champion impala in bigtop, im not objecting either, and maybe i can even help w/ the smoke tests and so on. On Mon, Dec 8, 2014 at 1:22 PM, Leidle, Rob <[email protected]> wrote: > Sorry, that should say “unlock the full feature set of the Hue > component”. > > From: <Leidle>, "Leidle, Rob" <[email protected]> > Date: Monday, December 8, 2014 at 10:21 AM > To: "[email protected]" <[email protected]> > Subject: Impala & Bigtop > > I was reading the recent thread about Impala with regards to Hue and I > was wondering why Impala has not been brought into Bigtop. This would > unlock the full feature set of the Hadoop component. Is there a technical > or philosophical reason? Or is it just something that hasn’t been done or > even considered yet? > -- jay vyas
