hadoop - hive 在分区表上添加列不起作用

Question

我分享了在分区 hive 表上添加列的经验。
如您所见，尽管有 CASCADE 功能，但 ALTER 还是让我的 table 刹车:(

在分区表上添加列

表说明

CREATE TABLE test (
a                       string,      
b                       string,
c                       string
)
PARTITIONED BY (
x                       string,
y                       string, 
z                       string
)
ROW FORMAT SERDE
'org.apache.hadoop.hive.ql.io.orc.OrcSerde'
STORED AS INPUTFORMAT
'org.apache.hadoop.hive.ql.io.orc.OrcInputFormat'
OUTPUTFORMAT
'org.apache.hadoop.hive.ql.io.orc.OrcOutputFormat'
TBLPROPERTIES (
'orc.compress'='SNAPPY'
);

复制表格

CREATE TABLE test_tmp...

hadoop distcp hdfs://.../test/* dfs://.../test_tmp

MSCK REPAIR TABLE test_tmp;

SELECT * FROM test_tmp
LIMIT 100

check : OK (I get results)

修改表格

ALTER TABLE test_tmp
ADD COLUMNS(
aa  timestamp,
bb  string,
cc  int,
dd  string
) CASCADE;

SELECT * FROM test_tmp
LIMIT 100

...
]], Vertex did not succeed due to OWN_TASK_FAILURE, failedTasks:1 killedTasks:19, Vertex vertex_1502459312997_187854_4_00 [Map 1] killed/failed due to:OWN_TASK_FAILURE]DAG did not succeed due to VERTEX_FAILURE. failedVertices:1 killedVertices:0
... 1 statement(s) executed, 0 rows affected, exec/fetch time: 21.655/0.000 sec  [0 successful, 1 errors]

check : KO (I get this error)

Answer 1

如果您使用的是 Hive 0.x 或 1.x，那么您可能是...的受害者

HIVE-10598   Vectorization borks when column is added to table.

...这是特定于 ORC 格式的，即使它在 JIRA 标签中并不明显。

从 Hive 2.0 开始有部分修复(即 ADD 已修复，但 DROP/RENAME/CHANGE 仍然残缺不全)感谢

HIVE-11981   ORC Schema Evolution Issues (Vectorized, ACID, and Non-Vectorized)

以及从 Hive 2.1.1 开始的另一个相关修复 CHANGE

HIVE-14355   Schema evolution for ORC in llap is broken for Int to String conversion

待续...