Hi Charles, That makes sense. I am a new contributor to Drill, so I'll start with a few small changes to get my feet wet. I will revisit implementing a storage plugin for elasticsearch in a few weeks once I am comfortable with the codebase. On another note, my main objective in contributing is to learn, so the more work that is involved in developing a feature, the better :)
Thanks, Badrul On Wed, Jul 24, 2019 at 7:09 PM Charles Givre <[email protected]> wrote: > Hi Badrul, > I think this page is seriously out of date. The copyright at the bottom > says 2014! Anyway, I would really like to see more storage plugins added > to Drill. With that said, it is not trivial to implement one. In my > opinion, an ElasticSearch plugin would be quite valuable. Regarding > ODBC/JDBC, you can connect TO Drill via ODBC and JDBC, however Drill cannot > connect to data sources via ODBC, only JDBC. > > In theory, there might be situations where you'd want to write a custom > storage plugin rather than using the generic JDBC connection. For > instance, you may have a system which performs much better via its own > interface rather than JDBC. I'm not 100% certain that Drill's JDBC storage > plugin supports pushdown projection as well. So there may be a use case > for it. But... it's a lot of work. > > -- C > > > > On Jul 24, 2019, at 6:53 PM, Badrul Chowdhury < > [email protected]> wrote: > > > > Hi, > > > > The list of suggested contributions here > > <https://drill.apache.org/docs/apache-drill-contribution-ideas/>includes > > writing a custom storage plugin for SQL databases. Drill already supports > > ODBC, so I was wondering why we would want to separately write custom > > plugins for different SQL backends? > > > > > > Thanks, > > Badrul > > -- Cheers, Badrul
