[
https://issues.apache.org/jira/browse/HIVE-6131?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13964556#comment-13964556
]
Pala M Muthaia commented on HIVE-6131:
--------------------------------------
Yes, one case where partition column is different than table column is the type
of an existing column is changed (e.g: from string to int) at the table level,
after a partition is created. So column1 is string on partition, but int on
table.
That is the cause of failure for the unit test
'org.apache.hadoop.hive.cli.TestCliDriver.testCliDriver_partition_wise_fileformat11'
you see above.
So we can't alter the partition schema to match table schema, because
underlying data in LazyBinaryColumnarSerde was written to adhere to old schema.
If you just change the schema, but not the data, the data will be read
incorrectly.
Yes serdes can handle additional columns by treating them as null valued.
However, we are trying to address a more complex case: e.g: Table and partition
has 3 columns. We add one more column at table level. Then at the partition, we
append (not overwrite) data with 4th column too. Now the partition has some
data with 3 columns, but some data with 4 columns. LazySimpleSerde can work in
this scenario, but LazyBinaryColumnarSerde doesn't.
> New columns after table alter result in null values despite data
> ----------------------------------------------------------------
>
> Key: HIVE-6131
> URL: https://issues.apache.org/jira/browse/HIVE-6131
> Project: Hive
> Issue Type: Bug
> Affects Versions: 0.11.0, 0.12.0, 0.13.0
> Reporter: James Vaughan
> Priority: Minor
> Attachments: HIVE-6131.1.patch
>
>
> Hi folks,
> I found and verified a bug on our CDH 4.0.3 install of Hive when adding
> columns to tables with Partitions using 'REPLACE COLUMNS'. I dug through the
> Jira a little bit and didn't see anything for it so hopefully this isn't just
> noise on the radar.
> Basically, when you alter a table with partitions and then reupload data to
> that partition, it doesn't seem to recognize the extra data that actually
> exists in HDFS- as in, returns NULL values on the new column despite having
> the data and recognizing the new column in the metadata.
> Here's some steps to reproduce using a basic table:
> 1. Run this hive command: CREATE TABLE jvaughan_test (col1 string)
> partitioned by (day string);
> 2. Create a simple file on the system with a couple of entries, something
> like "hi" and "hi2" separated by newlines.
> 3. Run this hive command, pointing it at the file: LOAD DATA LOCAL INPATH
> '<FILEDIR>' OVERWRITE INTO TABLE jvaughan_test PARTITION (day = '2014-01-02');
> 4. Confirm the data with: SELECT * FROM jvaughan_test WHERE day =
> '2014-01-02';
> 5. Alter the column definitions: ALTER TABLE jvaughan_test REPLACE COLUMNS
> (col1 string, col2 string);
> 6. Edit your file and add a second column using the default separator
> (ctrl+v, then ctrl+a in Vim) and add two more entries, such as "hi3" on the
> first row and "hi4" on the second
> 7. Run step 3 again
> 8. Check the data again like in step 4
> For me, this is the results that get returned:
> hive> select * from jvaughan_test where day = '2014-01-01';
> OK
> hi NULL 2014-01-02
> hi2 NULL 2014-01-02
> This is despite the fact that there is data in the file stored by the
> partition in HDFS.
> Let me know if you need any other information. The only workaround for me
> currently is to drop partitions for any I'm replacing data in and THEN
> reupload the new data file.
> Thanks,
> -James
--
This message was sent by Atlassian JIRA
(v6.2#6252)
