Apologies, it does indeed work when you add the correct JARs in Hive.
Tim
On Tue, Apr 17, 2012 at 3:33 PM, Tim Robertson wrote:
> Hi all,
>
> I am *really* interested in Hive-1634 (
> https://issues.apache.org/jira/browse/HIVE-1634). I have just built from
> Hive trunk using HBase 0.90.4 as the version (e.g. we run cdh3u2).
>
> We have an HBase table populated with Bytes, so I create the Hive table
> like so:
>
> CREATE EXTERNAL TABLE tim_hbase_occurrence (
> id int,
> scientific_name string,
> data_resource_id int
> ) STORED BY 'org.apache.hadoop.hive.hbase.HBaseStorageHandler' WITH
> SERDEPROPERTIES (
> "hbase.columns.mapping" = ":key,v:scientific_name,v:data_resource_id",
> "hbase.columns.storage.types" = "b,b,b"
> ) TBLPROPERTIES(
> "hbase.table.name" = "mini_occurrences",
> "hbase.table.default.storage.type" = "binary"
> );
>
> This suggests it understands the formats:
>
> hive> SELECT * FROM tim_hbase_occurrence LIMIT 3;
> OK
> 1444 Abies alba 1081
> 1445 Abies alba 1081
> 1446 Abies alba 1081
>
> But doing any queries, suggest not:
>
> hive> SELECT * FROM tim_hbase_occurrence WHERE scientific_name='Abies
> alba' limit 3;
> ...
> NULL Abies alba NULL
> NULL Abies alba NULL
> NULL Abies alba NULL
> Time taken: 9.668 seconds
>
> hive> SELECT * FROM tim_hbase_occurrence WHERE data_resource_id=1081;
> ...
> 0 (no records)
>
> Can anyone provide any guidance on this please?
>
> Thanks!
> Tim
>
>
