flink-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stefano Bortoli <s.bort...@gmail.com>
Subject Re: Oracle 11g number serialization: classcast problem
Date Wed, 23 Mar 2016 09:04:20 GMT
I had a look at the JDBC input format, and it does indeed interpret
BIGDECIMAL and NUMERIC values as double. The status of the JDBCInputFormat
is not adequate for real world use case, as for example does not deal with
NULL values.

However, with little effort we fixed few stuff and now we are getting to
something usable. We are actually trying to do something a-la sqoop,
therefore given a boundary query, we create the splits, and then assign it
to the input format to read the database with configurable parallelism. We
are still working on it. If we get to something stable and working, we'll
gladly share it.


2016-03-22 15:46 GMT+01:00 Chesnay Schepler <chesnay@apache.org>:

> The JDBC formats don't make any assumption as to what DB backend is used.
> A JDBC float in general is returned as a double, since that was the
> recommended mapping i found when i wrote the formats.
> Is the INT returned as a double as well?
> Note: The (runtime) output type is in no way connected to the TypeInfo you
> pass when constructing the format.
> On 21.03.2016 14:16, Stefano Bortoli wrote:
>> Hi squirrels,
>> I working on a flink job connecting to a Oracle DB. I started from the
>> JDBC example for Derby, and used the TupleTypeInfo to configure the fields
>> of the tuple as it is read.
>> The record of the example has 2 INT, 1 FLOAT and 2 VARCHAR. Apparently,
>> using Oracle, all the numbers are read as Double, causing a ClassCast
>> exception. Of course I can fix it by changing the TupleTypeInfo, but I
>> wonder whether there is some assumption for Oracle and Numbers.
>> Thanks a lot for your support!
>> saluti,
>> Stefano

View raw message