[IMPALA-3887] Planner tests failing due to metadata loading race with HDFS, fewer #hosts than expected

• Type: Bug
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: Impala 2.6.0, Impala 2.11.0
• Fix Version/s: Impala 2.12.0
• Component/s: Infrastructure
• Labels: broken-build flaky
• Assignee: Tianyi Wang
• Reporter: Jim Apple
• Created: 20/Jul/16 15:48
• Updated: 12/Jan/18 23:45
• Resolved: 12/Jan/18 00:04

Description

11:22:33.349 =================================== FAILURES ===================================
11:22:33.349  TestCorruptTableStats.test_compute_stats[exec_option: {'disable_codegen': False, 'abort_on_error': 1, 'exec_single_node_rows_threshold': 100, 'batch_size': 0, 'num_nodes': 0} | table_format: text/none] 
11:22:33.349 [gw0] linux2 -- Python 2.6.6 /data/jenkins/workspace/impala-umbrella-build-and-test/repos/Impala/bin/../infra/python/env/bin/python
11:22:33.349 metadata/test_compute_stats.py:40: in test_compute_stats
11:22:33.349     self.run_test_case('QueryTest/compute-stats', vector, unique_database)
11:22:33.349 common/impala_test_suite.py:323: in run_test_case
11:22:33.349     pytest.config.option.update_results)
11:22:33.349 common/test_result_verifier.py:357: in verify_raw_results
11:22:33.349     VERIFIER_MAP[verifier](expected, actual)
11:22:33.349 common/test_result_verifier.py:203: in verify_query_result_is_equal
11:22:33.349     assert expected_results == actual_results
11:22:33.349 E   assert Comparing QueryTestResults (expected vs actual):
11:22:33.349 E     100,3,'7.73KB','NOT CACHED','NOT CACHED','TEXT','false',regex:.* != 100,2,'7.73KB','NOT CACHED','NOT CACHED','TEXT','false','hdfs://localhost:20500/test-warehouse/test_compute_stats_42383876.db/alltypesnopart'

http://sandbox.jenkins.cloudera.com/job/impala-cdh5-2.6.0_5.8.0-exhaustive-release/57/consoleFull

Issue Links

• is duplicated by: IMPALA-6200 Flakiness in Planner Tests
• is related to: IMPALA-6394 GVO failed in "Waiting for HDFS replication"