Hive partition columns seem to prevent "select distinct" - distinct

Question

I have created a table in Hive like this:
CREATE TABLE application_path
(userId STRING, sessId BIGINT, accesstime BIGINT, actionId STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '#'
STORED AS TEXTFILE;
Running on this table the query:
SELECT DISTINCT userId FROM application_path;
gives the expected result:
user1#domain.com
user2#domain.com
user3#domain.com
...
Then I've changed the declaration to add a partition:
CREATE TABLE application_path
(sessId BIGINT, accesstime BIGINT, actionId STRING)
PARTITIONED BY(userId STRING)
ROW FORMAT DELIMITED
FIELDS TERMINATED BY '#'
STORED AS TEXTFILE;
Now the query SELECT DISTINCT userId... runs for seconds as before, but eventually returns anything.
I've just noticed the syntax:
SHOW PARTITIONS application_path;
but I was wondering if that's the only way to get unique (distinct) values from a partitioning column. The output of SHOW PARTITION is not even an exact replacement of what you would get from SELECT DISTINCT, since the column name is prefixed to each row:
hive> show partitions application_path;
OK
userid=user1#domain.com
userid=user2#domain.com
userid=user3#domain.com
...
What's strange to me is that usedId can be used in GROUP BY with other columns, like in:
SELECT userId, sessId FROM application_path GROUP BY userId, sessId;
but does return anything in:
SELECT userId FROM application_path GROUP BY userId;

Answer 1

I experienced the same issue, it will be fixed in 0.10
https://issues.apache.org/jira/browse/HIVE-2955