hive-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Prabath Abeysekara <>
Subject Query on writing a double value to Cassandra via Hive
Date Fri, 22 Feb 2013 04:27:00 GMT

First of all, pardon me if the $subject had already been discussed in this
mailing list.

I have a Column family created in Cassandra with comparator : Int32Type,
default_validation_class : DoubleType and key_validation_class : UTF8Type.
What I'm trying to do is to read a set of records from a source table (with
the same configurations as mentioned earlier) and write them back to the
aforementioned target column family via Hive. However, whenever I do this,
the target records get inserted to the target column family as bytes thus
becoming non-human readable when retrieved via the Cassandra JDBC driver as
pure double values. It's the same result I get when I do list the records
in the column family via the cassandra-cli tool as well. However, the
values properly get converted as doubles and can be viewed properly when
retrieved via Hive.

Is this the expected behaviour? If not, is there any other mechanism that
we can use to avoid the aforementioned behaviour?


View raw message