[ATLAS-2975] Hive hook generates duplicate column_lineage entities - ASF JIRA

Attach files

Attach Screenshot

Voters

Watch issue

Watchers

Create sub-task

Link

Clone

Update Comment Author

Replace String in Comment

Update Comment Visibility

Delete Comments

XML

Word

Printable

JSON

Details

Type: Bug
Status: Resolved
Priority: Major
Resolution: Fixed
Affects Version/s: 1.0.0, 0.8.3, 1.1.0
Fix Version/s: 0.8.4, 1.2.0, 2.0.0
Component/s: atlas-intg
Labels:
None

Description

Hive hook is expected to create one column-lineage entity for each column in the output table. However, for each output column, hive hook might generates multiple column-lineage entities when multiple partitions are involved - one entity for each partition. This can end up with large number of duplciate column-lineage entities, depending on the number of partitions. Such duplicate entities should be avoided.

Here is the sample HSQL to repro this issue:

CREATE TABLE visitors(name STRING, dob DATE) PARTITIONED BY (yob INT);
CREATE TABLE visitors_log(name STRING, dob DATE);

INSERT INTO TABLE visitors_log VALUES('John',  '1980-08-08'),
                                     ('Jack',  '1980-09-09'),
                                     ('Kevin', '1990-10-10'),
                                     ('Ken',   '1990-11-11'),
                                     ('Larry', '1995-12-12');

SET hive.exec.dynamic.partition.mode=nonstrict;

INSERT INTO TABLE visitors PARTITION(yob) SELECT name, dob, YEAR(dob) yob FROM visitors_log;

In above case, columns visitors.name and visitors.dob will have 3 input lineage - one for each partition 1980, 1990 and 1995.