Ok thanks for this info! Maybe this could be added to the documentation..what do you think?
Il Dom 2 Feb 2020, 08:37 Jingsong Li <jingsongl...@gmail.com> ha scritto: > Hi Flavio, > > You can use `JDBCTableSource`, and register it from > TableEnvionment.registerTableSource, you need provide > a OracleDialect, maybe just implement `canHandle` and > `defaultDriverName` is OK. > > Best, > Jingsong Lee > > On Sun, Feb 2, 2020 at 2:42 PM Jark Wu <imj...@gmail.com> wrote: > >> Hi Flavio, >> >> If you want to adjust the writing statement for Oracle, you can implement >> the JDBCDialect for Oracle, and pass to the JDBCUpsertTableSink when >> constructing via `JDBCOptions.Builder#setDialect`. In this way, you don't >> need to recompile the source code of flink-jdbc. >> >> Best, >> Jark >> >> On Fri, 31 Jan 2020 at 19:28, Flavio Pompermaier <pomperma...@okkam.it> >> wrote: >> >>> Hi to all, >>> I was looking at the Flink SQL API's and I discovered that only a few >>> drivers are supported [1], i.e. Mysql, Postgres and Derby. You could have >>> problems only on the writing side of the connector (TableSink) because you >>> need to adjust the override statement, but for the read part you shouldn't >>> have problems with dialects...am I wrong? >>> And what am I supposed to do right now if I want to connect to Oracle >>> using the Table API? Do I have to use the low level JDBCInputFormat? Is >>> there an easy way to connect to Oracle using the Table API without the need >>> to modify and recompile the source code of Flink (just adding some >>> interface implementation in the application JAR)? >>> >>> [1] >>> https://ci.apache.org/projects/flink/flink-docs-stable/dev/table/connect.html#jdbc-connector >>> >>> Best, >>> Flavio >>> >> > > -- > Best, Jingsong Lee >
