I have made some fixes to the JDBC storage plugin in order to make it work with an Oracle database. I can now obtain the table names. As stated in my previous email the JDBC API is not properly used in the case of an Oracle engine.
I'm now able to obtain the table names but I don't really understand how to construct the internal objects drill uses to keep track of the schemas and tables. Should I report this as a bug? Thank you jc On Wed, Dec 19, 2018 at 9:50 AM Jean-Claude Cote <jcc...@gmail.com> wrote: > I've configured drill to use an JDBC storage plugin. My connection string > is for an Oracle database. I have included the Oracle JDBC driver to my > drill deployment. > > The connection is established correctly. However the storage plugin fails > to retrieve the schema of the database. > > The JDBC API provides two ways to get metadata about the database: > getSchemas() and getCatalogs(). Depending on the database vendor one of > those calls will return the information. The storage plugin correctly tries > both calls. However I believe there is an issue when using Oracle. > > The JDBC API claims that getSchemas() returns a set of two columns "table > schema" and "table catalog" but in reality it just returns "table schema" > as a first column of the result set. I believe this to be a bug. Has anyone > tested the JDBC storage plugin against an Oracle DB and successfully > retrieved the list of schema and tables? > > This article explains in details the issue in retrieving schema > information using JDBC. > https://www.wisdomjobs.com/e-university/jdbc-tutorial-278/what-are-catalogs-and-schemas-1147.html > > Thank you > jc > >
