jprdonnelly/impala
1
0
Fork 0
mirror of https://github.com/apache/impala.git synced 2025-12-29 09:04:47 -05:00
Code Issues Projects Releases Wiki Activity

IMPALA-12409: Don't allow EXTERNAL Iceberg tables to point another Iceberg table in Hive catalog

Browse Source 
This patch forbids creating an EXTERNAL Iceberg table that points
to another Iceberg table in the Hive Catalog. I.e. the following should
be forbidden:

  CREATE EXTERNAL TABLE ice_ext
  STORED BY ICEBERG
  TBLPROPERTIES ('iceberg.table_identifier'='db.tbl');

Loading such tables should also raise an error. Users need to query
the original Iceberg tables. Alternatively they can create VIEWs if
they want to query tables with a different name.

Testing:
 * added e2e tests for CREATE EXTERNAL TABLE
 * added e2e test about loading such table

Change-Id: Ifb0d7f0e7ec40fba356bd58b43f68d070432de71
Reviewed-on: http://gerrit.cloudera.org:8080/20429
Reviewed-by: Impala Public Jenkins <impala-public-jenkins@cloudera.com>
Tested-by: Impala Public Jenkins <impala-public-jenkins@cloudera.com>
This commit is contained in:
Zoltan Borok-Nagy
2023-08-28 19:38:03 +02:00
committed by Impala Public Jenkins
parent c49f5d2778
commit a9aaeaa489
7 changed files with 123 additions and 124 deletions
Download Patch File Download Diff File Expand all files Collapse all files

27
tests/query_test/test_iceberg.py
View File

@@ -1086,6 +1086,33 @@ class TestIcebergTable(IcebergTestSuite):
assert parquet_column_name_type_list == iceberg_column_name_type_list
@SkipIfFS.hive
def test_hive_external_forbidden(self, vector, unique_database):
tbl_name = unique_database + ".hive_ext"
error_msg = ("cannot be loaded because it is an EXTERNAL table in the HiveCatalog "
"that points to another table. Query the original table instead.")
self.execute_query("create table {0} (i int) stored by iceberg".
format(tbl_name))
# 'iceberg.table_identifier' can refer to another table
self.run_stmt_in_hive("""alter table {0} set tblproperties
('external.table.purge'='false',
'iceberg.table_identifier'='functional_iceberg.iceberg_partitioned')""".
format(tbl_name))
ex = self.execute_query_expect_failure(self.client, "refresh {0}".format(tbl_name))
assert error_msg in str(ex)
# 'iceberg.mr.table.identifier' can refer to another table
self.run_stmt_in_hive("""
alter table {0} unset tblproperties('iceberg.table_identifier')""".
format(tbl_name))
self.run_stmt_in_hive("""alter table {0} set tblproperties
('iceberg.mr.table.identifier'='functional_iceberg.iceberg_partitioned')""".
format(tbl_name))
ex = self.execute_query_expect_failure(self.client, "refresh {0}".format(tbl_name))
assert error_msg in str(ex)
# 'name' can also refer to another table but cannot be set by Hive/Impala. Also,
# during table migration both Impala and Hive clears existing table properties
# See IMPALA-12410
@SkipIfFS.incorrent_reported_ec
def test_compute_stats(self, vector, unique_database):
self.run_test_case('QueryTest/iceberg-compute-stats', vector, unique_database)