This is an automated email from the ASF dual-hosted git repository.

michaelsmith pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/impala.git

commit 1b93c370b250211195a2b24ecf8c4e8c100dd679
Author: Riza Suminto <[email protected]>
AuthorDate: Thu Dec 14 11:55:32 2023 -0800

    IMPALA-12630: Deflake TestOrcStats.test_orc_stats
    
    After revert of IMPALA-9923 (b03e8ef95c856f499d17ea7815831e30e2e9f467),
    TestOrcStats.test_orc_stats becomes flaky in the following test case:
    
    select count(*) from tpch_orc_def.lineitem where l_orderkey = 1609411;
    
    This is because tpch_orc_def.lineitem sometimes loaded as 1 ORC file in
    one run, and loaded in as 4 ORC files in the other run (see JIRA
    comments for more analysis).
    
    This patch attempts to deflake the test by changing the test case to
    count over orders table instead. orders table is not as large as
    lineitem table, but more consistently written as 1 ORC file with more
    than 1 stripe.
    
    Testing:
    - Pass TestOrcStats.test_orc_stats after reloading tpch workload.
    
    Change-Id: Ibb7bb25b44878df3e037e750a985656ed6857a7d
    Reviewed-on: http://gerrit.cloudera.org:8080/20794
    Reviewed-by: Impala Public Jenkins <[email protected]>
    Tested-by: Impala Public Jenkins <[email protected]>
---
 .../workloads/functional-query/queries/QueryTest/orc-stats.test     | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git 
a/testdata/workloads/functional-query/queries/QueryTest/orc-stats.test 
b/testdata/workloads/functional-query/queries/QueryTest/orc-stats.test
index 14b14b00b..8a6a32f38 100644
--- a/testdata/workloads/functional-query/queries/QueryTest/orc-stats.test
+++ b/testdata/workloads/functional-query/queries/QueryTest/orc-stats.test
@@ -601,11 +601,11 @@ aggregation(SUM, RowsRead): 7300
 ---- QUERY
 # Test on a larger ORC file that has multiple stripes and each stripe has 
multiple row
 # groups.
-select count(*) from tpch_orc_def.lineitem where l_orderkey = 1609411;
+select count(*) from tpch_orc_def.orders where o_orderkey = 1;
 ---- RESULTS
-7
+1
 ---- RUNTIME_PROFILE
-aggregation(SUM, RowsRead): 13501
+aggregation(SUM, RowsRead): 10000
 ====
 ---- QUERY
 # Test pushing down IN-list predicate with literals outside the value range.

Reply via email to